Datasets:
vgainullin
/
xciting_data

Dataset card Data Studio Files
xet
 Community
1
paragraph_index
int64
sec
string
p_has_citation
int64
cites
string
citeids
list
pmid
int64
cited_id
string
sentences
string
all_sent_cites
list
sent_len
int64
sentence_batch_index
int64
sent_has_citation
float64
qc_fail
bool
cited_sentence
string
cites_in_sentence
list
cln_sentence
string
is_cap
bool
is_alpha
bool
ends_wp
bool
cit_qc
bool
lgtm
bool
__index_level_0__
int64
2
INTRODUCTION
1
10
[ "B10", "B11", "B12" ]
17,483,517
NA|pmid-2184437|NA
Thus, the reported value was a measure of the extent to which a particular alignment was better than background.
[ "10", "11", "12" ]
112
9,000
0
false
Thus, the reported value was a measure of the extent to which a particular alignment was better than background.
[]
Thus, the reported value was a measure of the extent to which a particular alignment was better than background.
true
true
true
true
true
1,432
3
INTRODUCTION
0
null
null
17,483,517
null
The use of methods such as this, which seek to obtain global or local optimal solutions to inference problems, is common in computational biology.
null
146
9,001
0
false
null
null
The use of methods such as this, which seek to obtain global or local optimal solutions to inference problems, is common in computational biology.
true
true
true
true
true
1,433
3
INTRODUCTION
0
null
null
17,483,517
null
Typically, however, the probability of even the best arrangement of motif sites is extremely small.
null
99
9,002
0
false
null
null
Typically, however, the probability of even the best arrangement of motif sites is extremely small.
true
true
true
true
true
1,433
3
INTRODUCTION
0
null
null
17,483,517
null
That is, since motif detection is a high-dimensional problem, from a Bayesian viewpoint, the data likelihood will contain an immense number of terms, of which the optimal solution is simply one.
null
194
9,003
0
false
null
null
That is, since motif detection is a high-dimensional problem, from a Bayesian viewpoint, the data likelihood will contain an immense number of terms, of which the optimal solution is simply one.
true
true
true
true
true
1,433
3
INTRODUCTION
0
null
null
17,483,517
null
From this perspective, the question arises, ‘How representative is the optimum when its probability is very small compared to the overall probability mass?’
null
156
9,004
0
false
null
null
From this perspective, the question arises, ‘How representative is the optimum when its probability is very small compared to the overall probability mass?’
true
true
false
true
false
1,433
4
INTRODUCTION
1
13
[ "B13", "B2", "B14", "B13", "B15", "B2" ]
17,483,517
pmid-16043502|NA|NA|pmid-16043502|pmid-16631786|NA
It has been shown in RNA secondary structure prediction (13) and TFBS discovery algorithms (2,14) that reliance on the optimal solution can be misleading and can adversely affect prediction accuracy.
[ "13", "2", "14", "13", "15", "2" ]
199
9,005
1
false
It has been shown in RNA secondary structure prediction and TFBS discovery algorithms that reliance on the optimal solution can be misleading and can adversely affect prediction accuracy.
[ "13", "2,14" ]
It has been shown in RNA secondary structure prediction and TFBS discovery algorithms that reliance on the optimal solution can be misleading and can adversely affect prediction accuracy.
true
true
true
true
true
1,434
4
INTRODUCTION
1
13
[ "B13", "B2", "B14", "B13", "B15", "B2" ]
17,483,517
pmid-16043502|NA|NA|pmid-16043502|pmid-16631786|NA
Specifically, Ding et al.
[ "13", "2", "14", "13", "15", "2" ]
25
9,006
0
false
Specifically, Ding et al.
[]
Specifically, Ding et al.
true
true
true
true
true
1,434
4
INTRODUCTION
1
13
[ "B13", "B2", "B14", "B13", "B15", "B2" ]
17,483,517
pmid-16043502|NA|NA|pmid-16043502|pmid-16631786|NA
(13,15) showed that centroid estimates reduced errors in RNA secondary structure prediction by 30%, while simultaneously improving sensitivity, and Newberg et al.
[ "13", "2", "14", "13", "15", "2" ]
162
9,007
0
false
showed that centroid estimates reduced errors in RNA secondary structure prediction by 30%, while simultaneously improving sensitivity, and Newberg et al.
[ "13,15" ]
showed that centroid estimates reduced errors in RNA secondary structure prediction by 30%, while simultaneously improving sensitivity, and Newberg et al.
false
true
true
true
false
1,434
4
INTRODUCTION
1
2
[ "B13", "B2", "B14", "B13", "B15", "B2" ]
17,483,517
pmid-16043502|NA|NA|pmid-16043502|pmid-16631786|NA
(2) showed similar substantial improvements over algorithms finding local optima for TFBS discovery in sequences from phylogenetically closely related species.
[ "13", "2", "14", "13", "15", "2" ]
159
9,008
1
false
showed similar substantial improvements over algorithms finding local optima for TFBS discovery in sequences from phylogenetically closely related species.
[ "2" ]
showed similar substantial improvements over algorithms finding local optima for TFBS discovery in sequences from phylogenetically closely related species.
false
true
true
true
false
1,434
4
INTRODUCTION
1
13
[ "B13", "B2", "B14", "B13", "B15", "B2" ]
17,483,517
pmid-16043502|NA|NA|pmid-16043502|pmid-16631786|NA
Centroid solutions garner information from the full ensemble of solutions, while MAP solutions focus exclusively on the single most probable point.
[ "13", "2", "14", "13", "15", "2" ]
147
9,009
0
false
Centroid solutions garner information from the full ensemble of solutions, while MAP solutions focus exclusively on the single most probable point.
[]
Centroid solutions garner information from the full ensemble of solutions, while MAP solutions focus exclusively on the single most probable point.
true
true
true
true
true
1,434
5
INTRODUCTION
1
12
[ "B12", "B3" ]
17,483,517
NA|pmid-12824370
The user supplies to the algorithm a collection of sequences in FASTA format and enters several parameters, such as motif widths, as described below.
[ "12", "3" ]
149
9,010
0
false
The user supplies to the algorithm a collection of sequences in FASTA format and enters several parameters, such as motif widths, as described below.
[]
The user supplies to the algorithm a collection of sequences in FASTA format and enters several parameters, such as motif widths, as described below.
true
true
true
true
true
1,435
5
INTRODUCTION
1
12
[ "B12", "B3" ]
17,483,517
NA|pmid-12824370
The centroid algorithm begins in a manner similar to previous Gibbs sampling algorithms.
[ "12", "3" ]
88
9,011
0
false
The centroid algorithm begins in a manner similar to previous Gibbs sampling algorithms.
[]
The centroid algorithm begins in a manner similar to previous Gibbs sampling algorithms.
true
true
true
true
true
1,435
5
INTRODUCTION
1
12
[ "B12", "B3" ]
17,483,517
NA|pmid-12824370
It is initialized with a, typically random, alignment.
[ "12", "3" ]
54
9,012
0
false
It is initialized with a, typically random, alignment.
[]
It is initialized with a, typically random, alignment.
true
true
true
true
true
1,435
5
INTRODUCTION
1
12
[ "B12", "B3" ]
17,483,517
NA|pmid-12824370
From this alignment, motif models are calculated (12).
[ "12", "3" ]
54
9,013
1
false
From this alignment, motif models are calculated.
[ "12" ]
From this alignment, motif models are calculated.
true
true
true
true
true
1,435
5
INTRODUCTION
1
12
[ "B12", "B3" ]
17,483,517
NA|pmid-12824370
The sampling procedure then proceeds through the following steps: A sequence is selected, and the probability of each possible number of sites, up to the maximum specified by the user, is calculated based on the current model;the number of sites is sampled;the predicted positions and types of the sites are sampled base...
[ "12", "3" ]
388
9,014
0
false
The sampling procedure then proceeds through the following steps: A sequence is selected, and the probability of each possible number of sites, up to the maximum specified by the user, is calculated based on the current model;the number of sites is sampled;the predicted positions and types of the sites are sampled base...
[]
The sampling procedure then proceeds through the following steps: A sequence is selected, and the probability of each possible number of sites, up to the maximum specified by the user, is calculated based on the current model;the number of sites is sampled;the predicted positions and types of the sites are sampled base...
true
true
true
true
true
1,435
5
INTRODUCTION
1
3
[ "B12", "B3" ]
17,483,517
NA|pmid-12824370
(3);the motif models are updated based on the sampled sites in all sequences.
[ "12", "3" ]
77
9,015
1
false
;the motif models are updated based on the sampled sites in all sequences.
[ "3" ]
;the motif models are updated based on the sampled sites in all sequences.
false
false
true
true
false
1,435
6
INTRODUCTION
0
null
null
17,483,517
null
A sequence is selected, and the probability of each possible number of sites, up to the maximum specified by the user, is calculated based on the current model;
null
160
9,016
0
false
null
null
A sequence is selected, and the probability of each possible number of sites, up to the maximum specified by the user, is calculated based on the current model;
true
true
false
true
false
1,436
7
INTRODUCTION
0
null
null
17,483,517
null
the number of sites is sampled;
null
31
9,017
0
false
null
null
the number of sites is sampled;
false
true
false
true
false
1,437
8
INTRODUCTION
1
3
[ "B3" ]
17,483,517
pmid-12824370
the predicted positions and types of the sites are sampled based on their probabilities, calculated as described by Thompson et al.
[ "3" ]
131
9,018
0
false
the predicted positions and types of the sites are sampled based on their probabilities, calculated as described by Thompson et al.
[]
the predicted positions and types of the sites are sampled based on their probabilities, calculated as described by Thompson et al.
false
true
true
true
false
1,438
9
INTRODUCTION
0
null
null
17,483,517
null
the motif models are updated based on the sampled sites in all sequences.
null
73
9,019
0
false
null
null
the motif models are updated based on the sampled sites in all sequences.
false
true
true
true
false
1,439
10
INTRODUCTION
0
null
null
17,483,517
null
An iteration of the algorithm consists of the completion of Steps 1–4 for each sequence.
null
88
9,020
0
false
null
null
An iteration of the algorithm consists of the completion of Steps 1–4 for each sequence.
true
true
true
true
true
1,440
10
INTRODUCTION
0
null
null
17,483,517
null
In previous versions, this process repeated until the MAP failed to increase for a fixed number of iterations.
null
110
9,021
0
false
null
null
In previous versions, this process repeated until the MAP failed to increase for a fixed number of iterations.
true
true
true
true
true
1,440
10
INTRODUCTION
0
null
null
17,483,517
null
To obtain a sampling solution, we allow the algorithm to repeat the above procedure through a burn-in period, typically 2000 iterations.
null
136
9,022
0
false
null
null
To obtain a sampling solution, we allow the algorithm to repeat the above procedure through a burn-in period, typically 2000 iterations.
true
true
true
true
true
1,440
10
INTRODUCTION
0
null
null
17,483,517
null
The burn-in period is required for the sampler to move away from transient effects of the particular initial conditions.
null
120
9,023
0
false
null
null
The burn-in period is required for the sampler to move away from transient effects of the particular initial conditions.
true
true
true
true
true
1,440
10
INTRODUCTION
0
null
null
17,483,517
null
After the burn-in period, the sampler proceeds, again through a fixed number of iterations (typically 8000).
null
108
9,024
0
false
null
null
After the burn-in period, the sampler proceeds, again through a fixed number of iterations (typically 8000).
true
true
true
true
true
1,440
10
INTRODUCTION
0
null
null
17,483,517
null
During this sampling process, the algorithm tracks each sampled position.
null
73
9,025
0
false
null
null
During this sampling process, the algorithm tracks each sampled position.
true
true
true
true
true
1,440
10
INTRODUCTION
0
null
null
17,483,517
null
The entire process (burn-in and sampling iterations) is repeated with a number of different random starting alignments called ‘seeds’.
null
134
9,026
0
false
null
null
The entire process (burn-in and sampling iterations) is repeated with a number of different random starting alignments called ‘seeds’.
true
true
true
true
true
1,440
10
INTRODUCTION
0
null
null
17,483,517
null
By default, 20 seeds are used.
null
30
9,027
0
false
null
null
By default, 20 seeds are used.
true
true
true
true
true
1,440
10
INTRODUCTION
0
null
null
17,483,517
null
The samples from each seed are accumulated, and a centroid alignment solution is obtained from the accumulated samples; the centroid is the alignment that minimizes the sum of the pair-wise distances between it and each of the alignments in the collection.
null
256
9,028
0
false
null
null
The samples from each seed are accumulated, and a centroid alignment solution is obtained from the accumulated samples; the centroid is the alignment that minimizes the sum of the pair-wise distances between it and each of the alignments in the collection.
true
true
true
true
true
1,440
10
INTRODUCTION
0
null
null
17,483,517
null
Thus, the centroid is defined in terms of a distance measure between pairs of proposed alignments.
null
98
9,029
0
false
null
null
Thus, the centroid is defined in terms of a distance measure between pairs of proposed alignments.
true
true
true
true
true
1,440
10
INTRODUCTION
0
null
null
17,483,517
null
The centroid alignment is calculated via a dynamic programming algorithm.
null
73
9,030
0
false
null
null
The centroid alignment is calculated via a dynamic programming algorithm.
true
true
true
true
true
1,440
11
INTRODUCTION
1
12
[ "B12", "B2" ]
17,483,517
NA|NA
In previous versions of the sampler, the model update step (Step 4 above) was accomplished using the predictive update method (12).
[ "12", "2" ]
131
9,031
1
false
In previous versions of the sampler, the model update step (Step 4 above) was accomplished using the predictive update method.
[ "12" ]
In previous versions of the sampler, the model update step (Step 4 above) was accomplished using the predictive update method.
true
true
true
true
true
1,441
11
INTRODUCTION
1
12
[ "B12", "B2" ]
17,483,517
NA|NA
The centroid sampler performs the model update step by sampling a new model from the posterior Dirichlet distribution of motif or background models.
[ "12", "2" ]
148
9,032
0
false
The centroid sampler performs the model update step by sampling a new model from the posterior Dirichlet distribution of motif or background models.
[]
The centroid sampler performs the model update step by sampling a new model from the posterior Dirichlet distribution of motif or background models.
true
true
true
true
true
1,441
11
INTRODUCTION
1
12
[ "B12", "B2" ]
17,483,517
NA|NA
Starting with the existing model Θ, the algorithm draws a new model, Θp, using the motif or background counts from Dir (c + β), where Dir is the Dirichlet distribution, and c and β are the current count and pseudo-count vectors.
[ "12", "2" ]
228
9,033
0
false
Starting with the existing model Θ, the algorithm draws a new model, Θp, using the motif or background counts from Dir (c + β), where Dir is the Dirichlet distribution, and c and β are the current count and pseudo-count vectors.
[]
Starting with the existing model Θ, the algorithm draws a new model, Θp, using the motif or background counts from Dir (c + β), where Dir is the Dirichlet distribution, and c and β are the current count and pseudo-count vectors.
true
true
true
true
true
1,441
11
INTRODUCTION
1
12
[ "B12", "B2" ]
17,483,517
NA|NA
While predictive update works when at most one new binding site is chosen between motif model updates, it is not entirely appropriate in the present context, where multiple binding sites are chosen between model updates.
[ "12", "2" ]
220
9,034
0
false
While predictive update works when at most one new binding site is chosen between motif model updates, it is not entirely appropriate in the present context, where multiple binding sites are chosen between model updates.
[]
While predictive update works when at most one new binding site is chosen between motif model updates, it is not entirely appropriate in the present context, where multiple binding sites are chosen between model updates.
true
true
true
true
true
1,441
11
INTRODUCTION
1
2
[ "B12", "B2" ]
17,483,517
NA|NA
This new model update method is of greatest value in the identification of sites among aligned sequences derived from multiple phylogenetically related species (2).
[ "12", "2" ]
164
9,035
1
false
This new model update method is of greatest value in the identification of sites among aligned sequences derived from multiple phylogenetically related species.
[ "2" ]
This new model update method is of greatest value in the identification of sites among aligned sequences derived from multiple phylogenetically related species.
true
true
true
true
true
1,441
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
The Gibbs Sampler Web site consists of three layers, each offering an increasing number of options for control of the sampling process.
[ "16", "2", "12", "16" ]
135
9,036
0
false
The Gibbs Sampler Web site consists of three layers, each offering an increasing number of options for control of the sampling process.
[]
The Gibbs Sampler Web site consists of three layers, each offering an increasing number of options for control of the sampling process.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
The first page, shown in Figure 1, allows the user to input sequences, select the version of the Gibbs Sampler, and control the basic motif parameters (16).
[ "16", "2", "12", "16" ]
156
9,037
1
false
The first page, shown in Figure 1, allows the user to input sequences, select the version of the Gibbs Sampler, and control the basic motif parameters.
[ "16" ]
The first page, shown in Figure 1, allows the user to input sequences, select the version of the Gibbs Sampler, and control the basic motif parameters.
true
true
true
true
true
1,442
12
INTRODUCTION
1
2
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
While we continue to make earlier versions available for selection on this page, in most circumstances the centroid sampler should return better results (2).
[ "16", "2", "12", "16" ]
157
9,038
1
false
While we continue to make earlier versions available for selection on this page, in most circumstances the centroid sampler should return better results.
[ "2" ]
While we continue to make earlier versions available for selection on this page, in most circumstances the centroid sampler should return better results.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
An e-mail address, a set of sequences in FASTA format, an optional initial guess of the total number of sites, the number of conserved positions in the motif sites, and the maximum allowable number of sites in any one sequence are entered on this page.
[ "16", "2", "12", "16" ]
252
9,039
0
false
An e-mail address, a set of sequences in FASTA format, an optional initial guess of the total number of sites, the number of conserved positions in the motif sites, and the maximum allowable number of sites in any one sequence are entered on this page.
[]
An e-mail address, a set of sequences in FASTA format, an optional initial guess of the total number of sites, the number of conserved positions in the motif sites, and the maximum allowable number of sites in any one sequence are entered on this page.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
The estimate of the number of sites affects the initial starting solution for the burn-in process.
[ "16", "2", "12", "16" ]
98
9,040
0
false
The estimate of the number of sites affects the initial starting solution for the burn-in process.
[]
The estimate of the number of sites affects the initial starting solution for the burn-in process.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
If it is not supplied, the default of one site for each motif type for each sequence is used.
[ "16", "2", "12", "16" ]
93
9,041
0
false
If it is not supplied, the default of one site for each motif type for each sequence is used.
[]
If it is not supplied, the default of one site for each motif type for each sequence is used.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
We have found this default adequate for most datasets, and the centroid sampler is relatively insensitive to reasonably small changes in this value.
[ "16", "2", "12", "16" ]
148
9,042
0
false
We have found this default adequate for most datasets, and the centroid sampler is relatively insensitive to reasonably small changes in this value.
[]
We have found this default adequate for most datasets, and the centroid sampler is relatively insensitive to reasonably small changes in this value.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
The number of conserved positions in the motif model(s) is a required parameter.
[ "16", "2", "12", "16" ]
80
9,043
0
false
The number of conserved positions in the motif model(s) is a required parameter.
[]
The number of conserved positions in the motif model(s) is a required parameter.
true
true
true
true
true
1,442
12
INTRODUCTION
1
12
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
This value sets the minimum width of the predicted sites, although sites may fragment to a greater width by the inclusion of non-conserved positions (12).
[ "16", "2", "12", "16" ]
154
9,044
1
false
This value sets the minimum width of the predicted sites, although sites may fragment to a greater width by the inclusion of non-conserved positions.
[ "12" ]
This value sets the minimum width of the predicted sites, although sites may fragment to a greater width by the inclusion of non-conserved positions.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
Motif widths for multiple models can be entered, although it is best to use no more motif models than is reasonable given the number of expected TFBS types.
[ "16", "2", "12", "16" ]
156
9,045
0
false
Motif widths for multiple models can be entered, although it is best to use no more motif models than is reasonable given the number of expected TFBS types.
[]
Motif widths for multiple models can be entered, although it is best to use no more motif models than is reasonable given the number of expected TFBS types.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
Increasing the number of motif models beyond the number of relevant site types should not adversely affect the solutions, if the number of burn-in and sample iterations is adequate (described below), because extra models will not sample sites sufficiently to be included in the centroid.
[ "16", "2", "12", "16" ]
287
9,046
0
false
Increasing the number of motif models beyond the number of relevant site types should not adversely affect the solutions, if the number of burn-in and sample iterations is adequate (described below), because extra models will not sample sites sufficiently to be included in the centroid.
[]
Increasing the number of motif models beyond the number of relevant site types should not adversely affect the solutions, if the number of burn-in and sample iterations is adequate (described below), because extra models will not sample sites sufficiently to be included in the centroid.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
However, as the number of models increases, the program runtime increases (described below).
[ "16", "2", "12", "16" ]
92
9,047
0
false
However, as the number of models increases, the program runtime increases (described below).
[]
However, as the number of models increases, the program runtime increases (described below).
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
The maximum number of sites in a single sequence is also a required parameter for the centroid sampler.
[ "16", "2", "12", "16" ]
103
9,048
0
false
The maximum number of sites in a single sequence is also a required parameter for the centroid sampler.
[]
The maximum number of sites in a single sequence is also a required parameter for the centroid sampler.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
The value entered for this parameter should be based on knowledge of the biological system under study.
[ "16", "2", "12", "16" ]
103
9,049
0
false
The value entered for this parameter should be based on knowledge of the biological system under study.
[]
The value entered for this parameter should be based on knowledge of the biological system under study.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
For example, when analyzing bacterial intergenic sequences for TFBSs, a value of two or three is typically used, whereas for eukaryotic data, this number is typically set higher.
[ "16", "2", "12", "16" ]
178
9,050
0
false
For example, when analyzing bacterial intergenic sequences for TFBSs, a value of two or three is typically used, whereas for eukaryotic data, this number is typically set higher.
[]
For example, when analyzing bacterial intergenic sequences for TFBSs, a value of two or three is typically used, whereas for eukaryotic data, this number is typically set higher.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
This parameter sets the maximum for the total sum of all motif sites in any one sequence.
[ "16", "2", "12", "16" ]
89
9,051
0
false
This parameter sets the maximum for the total sum of all motif sites in any one sequence.
[]
This parameter sets the maximum for the total sum of all motif sites in any one sequence.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
The sequence data can be pasted into the entry window or uploaded from a file.
[ "16", "2", "12", "16" ]
78
9,052
0
false
The sequence data can be pasted into the entry window or uploaded from a file.
[]
The sequence data can be pasted into the entry window or uploaded from a file.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
Each entry field has an associated hyperlink, which leads to a page describing the required data format.
[ "16", "2", "12", "16" ]
104
9,053
0
false
Each entry field has an associated hyperlink, which leads to a page describing the required data format.
[]
Each entry field has an associated hyperlink, which leads to a page describing the required data format.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
From this entry screen, default options will be automatically selected for the sampling parameters.
[ "16", "2", "12", "16" ]
99
9,054
0
false
From this entry screen, default options will be automatically selected for the sampling parameters.
[]
From this entry screen, default options will be automatically selected for the sampling parameters.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
The defaults for the centroid sampler include the use of a heterogeneous background model (16), 20 random seeds, a burn-in period of 2000 iterations and a sampling period of 8000 iterations.
[ "16", "2", "12", "16" ]
190
9,055
1
false
The defaults for the centroid sampler include the use of a heterogeneous background model, 20 random seeds, a burn-in period of 2000 iterations and a sampling period of 8000 iterations.
[ "16" ]
The defaults for the centroid sampler include the use of a heterogeneous background model, 20 random seeds, a burn-in period of 2000 iterations and a sampling period of 8000 iterations.
true
true
true
true
true
1,442
12
INTRODUCTION
1
16
[ "B16", "B2", "B12", "B16" ]
17,483,517
pmid-10068691|NA|NA|pmid-10068691
Figure 1.The basic Gibbs Centroid Sampler entry screen.
[ "16", "2", "12", "16" ]
55
9,056
0
false
Figure 1.The basic Gibbs Centroid Sampler entry screen.
[]
Figure 1.The basic Gibbs Centroid Sampler entry screen.
true
true
true
true
true
1,442
13
INTRODUCTION
0
null
null
17,483,517
null
The basic Gibbs Centroid Sampler entry screen.
null
46
9,057
0
false
null
null
The basic Gibbs Centroid Sampler entry screen.
true
true
true
true
true
1,443
14
INTRODUCTION
1
3
[ "B3", "B9" ]
17,483,517
pmid-12824370|NA
Selection of the ‘Show Advanced Options’ link opens a page with several more options (Figure 2).
[ "3", "9" ]
96
9,058
0
false
Selection of the ‘Show Advanced Options’ link opens a page with several more options (Figure 2).
[]
Selection of the ‘Show Advanced Options’ link opens a page with several more options (Figure 2).
true
true
true
true
true
1,444
14
INTRODUCTION
1
3
[ "B3", "B9" ]
17,483,517
pmid-12824370|NA
Most of these, such as options for palindromic models, fragmentation, the Wilcoxon signed-rank test and the number of random seeds, are available for all sampling modes (site, motif, recursive and centroid) and have been described earlier (3,9) New options for controlling the behavior of the centroid sampler are now al...
[ "3", "9" ]
346
9,059
0
false
Most of these, such as options for palindromic models, fragmentation, the Wilcoxon signed-rank test and the number of random seeds, are available for all sampling modes (site, motif, recursive and centroid) and have been described earlier New options for controlling the behavior of the centroid sampler are now also pre...
[ "3,9" ]
Most of these, such as options for palindromic models, fragmentation, the Wilcoxon signed-rank test and the number of random seeds, are available for all sampling modes (site, motif, recursive and centroid) and have been described earlier New options for controlling the behavior of the centroid sampler are now also pre...
true
true
true
true
true
1,444
14
INTRODUCTION
1
3
[ "B3", "B9" ]
17,483,517
pmid-12824370|NA
The ‘Burn-in Period’ and ‘Samples’ fields control the numbers of burn-in and sampling iterations for each seed; these fields are disabled when non-centroid sampling modes are selected.
[ "3", "9" ]
184
9,060
0
false
The ‘Burn-in Period’ and ‘Samples’ fields control the numbers of burn-in and sampling iterations for each seed; these fields are disabled when non-centroid sampling modes are selected.
[]
The ‘Burn-in Period’ and ‘Samples’ fields control the numbers of burn-in and sampling iterations for each seed; these fields are disabled when non-centroid sampling modes are selected.
true
true
true
true
true
1,444
14
INTRODUCTION
1
3
[ "B3", "B9" ]
17,483,517
pmid-12824370|NA
Initially, when the centroid sampler is selected, the ‘Burn-in Period’ and ‘Samples’ fields contain default values.
[ "3", "9" ]
115
9,061
0
false
Initially, when the centroid sampler is selected, the ‘Burn-in Period’ and ‘Samples’ fields contain default values.
[]
Initially, when the centroid sampler is selected, the ‘Burn-in Period’ and ‘Samples’ fields contain default values.
true
true
true
true
true
1,444
14
INTRODUCTION
1
3
[ "B3", "B9" ]
17,483,517
pmid-12824370|NA
We have found the defaults of 2000 iterations for burn-in and 8000 sampling iterations to be broadly applicable for prokaryotic or eukaryotic data of modest size.
[ "3", "9" ]
162
9,062
0
false
We have found the defaults of 2000 iterations for burn-in and 8000 sampling iterations to be broadly applicable for prokaryotic or eukaryotic data of modest size.
[]
We have found the defaults of 2000 iterations for burn-in and 8000 sampling iterations to be broadly applicable for prokaryotic or eukaryotic data of modest size.
true
true
true
true
true
1,444
14
INTRODUCTION
1
3
[ "B3", "B9" ]
17,483,517
pmid-12824370|NA
However, for small datasets, in the order of 10 to 20 sequences, each of <200 nucleotides, our experience has shown that the burn-in and sample iterations can be reduced (to 1000 and 4000, respectively) without adversely affecting the results.
[ "3", "9" ]
243
9,063
0
false
However, for small datasets, in the order of 10 to 20 sequences, each of <200 nucleotides, our experience has shown that the burn-in and sample iterations can be reduced (to 1000 and 4000, respectively) without adversely affecting the results.
[]
However, for small datasets, in the order of 10 to 20 sequences, each of <200 nucleotides, our experience has shown that the burn-in and sample iterations can be reduced (to 1000 and 4000, respectively) without adversely affecting the results.
true
true
true
true
true
1,444
14
INTRODUCTION
1
3
[ "B3", "B9" ]
17,483,517
pmid-12824370|NA
Conversely, for large datasets (>50 sequences, each of 5000 to 10,000 nucleotides) where the TFBS are likely short and not well conserved, as is common in eukaryotic sequences, the number of iterations should be increased for both parameters.
[ "3", "9" ]
242
9,064
0
false
Conversely, for large datasets (>50 sequences, each of 5000 to 10,000 nucleotides) where the TFBS are likely short and not well conserved, as is common in eukaryotic sequences, the number of iterations should be increased for both parameters.
[]
Conversely, for large datasets (>50 sequences, each of 5000 to 10,000 nucleotides) where the TFBS are likely short and not well conserved, as is common in eukaryotic sequences, the number of iterations should be increased for both parameters.
true
true
true
true
true
1,444
14
INTRODUCTION
1
3
[ "B3", "B9" ]
17,483,517
pmid-12824370|NA
Figure 2.The Gibbs Centroid ‘Advanced options’ entry page.
[ "3", "9" ]
58
9,065
0
false
Figure 2.The Gibbs Centroid ‘Advanced options’ entry page.
[]
Figure 2.The Gibbs Centroid ‘Advanced options’ entry page.
true
true
true
true
true
1,444
15
INTRODUCTION
0
null
null
17,483,517
null
The Gibbs Centroid ‘Advanced options’ entry page.
null
49
9,066
0
false
null
null
The Gibbs Centroid ‘Advanced options’ entry page.
true
true
true
true
true
1,445
16
INTRODUCTION
1
17
[ "B17", "B12" ]
17,483,517
NA|NA
It is important to note a difficulty that can arise when the centroid sampler is used with multiple motif models; specifically, the non-indentifiability of models from finite mixtures, stemming from label switching (17) among the various restarts of the algorithm.
[ "17", "12" ]
264
9,067
1
false
It is important to note a difficulty that can arise when the centroid sampler is used with multiple motif models; specifically, the non-indentifiability of models from finite mixtures, stemming from label switching among the various restarts of the algorithm.
[ "17" ]
It is important to note a difficulty that can arise when the centroid sampler is used with multiple motif models; specifically, the non-indentifiability of models from finite mixtures, stemming from label switching among the various restarts of the algorithm.
true
true
true
true
true
1,446
16
INTRODUCTION
1
17
[ "B17", "B12" ]
17,483,517
NA|NA
Gibbs sampling is inherently a stochastic procedure; in order to avoid being trapped in regions of low probability, the sampling process is restarted a number of different times with different starting seeds.
[ "17", "12" ]
208
9,068
0
false
Gibbs sampling is inherently a stochastic procedure; in order to avoid being trapped in regions of low probability, the sampling process is restarted a number of different times with different starting seeds.
[]
Gibbs sampling is inherently a stochastic procedure; in order to avoid being trapped in regions of low probability, the sampling process is restarted a number of different times with different starting seeds.
true
true
true
true
true
1,446
16
INTRODUCTION
1
17
[ "B17", "B12" ]
17,483,517
NA|NA
When multiple motif models are used, the separate seeds can converge to similar solutions, with different orderings of the motif models.
[ "17", "12" ]
136
9,069
0
false
When multiple motif models are used, the separate seeds can converge to similar solutions, with different orderings of the motif models.
[]
When multiple motif models are used, the separate seeds can converge to similar solutions, with different orderings of the motif models.
true
true
true
true
true
1,446
16
INTRODUCTION
1
17
[ "B17", "B12" ]
17,483,517
NA|NA
For example, in the case of two motif models, a particular seed may converge to a set of sites for model A and sites for model B.
[ "17", "12" ]
129
9,070
0
false
For example, in the case of two motif models, a particular seed may converge to a set of sites for model A and sites for model B.
[]
For example, in the case of two motif models, a particular seed may converge to a set of sites for model A and sites for model B.
true
true
true
true
true
1,446
16
INTRODUCTION
1
17
[ "B17", "B12" ]
17,483,517
NA|NA
Another seed may converge to the same overall collection of sites, but with the sites previously labeled as model A now labeled as model B, and sites previously labeled as model B now labeled as model A.
[ "17", "12" ]
203
9,071
0
false
Another seed may converge to the same overall collection of sites, but with the sites previously labeled as model A now labeled as model B, and sites previously labeled as model B now labeled as model A.
[]
Another seed may converge to the same overall collection of sites, but with the sites previously labeled as model A now labeled as model B, and sites previously labeled as model B now labeled as model A.
true
true
true
true
true
1,446
16
INTRODUCTION
1
17
[ "B17", "B12" ]
17,483,517
NA|NA
The centroid solution is obtained by summing the number of times a given position (i.e.
[ "17", "12" ]
87
9,072
0
false
The centroid solution is obtained by summing the number of times a given position (i.e.
[]
The centroid solution is obtained by summing the number of times a given position (i.e.
true
true
true
true
true
1,446
16
INTRODUCTION
1
17
[ "B17", "B12" ]
17,483,517
NA|NA
site) is sampled across all restarts and models, which means that sites from multiple models are not separated in the output.
[ "17", "12" ]
125
9,073
0
false
site) is sampled across all restarts and models, which means that sites from multiple models are not separated in the output.
[]
site) is sampled across all restarts and models, which means that sites from multiple models are not separated in the output.
false
true
true
true
false
1,446
16
INTRODUCTION
1
12
[ "B17", "B12" ]
17,483,517
NA|NA
Furthermore, different fragmentation models (12) can be generated among the different seed runs, giving rise to a collection of centroid sites that differ in length, and making it difficult to visualize the TFBSs in a more traditional probability matrix representation.
[ "17", "12" ]
269
9,074
1
false
Furthermore, different fragmentation models can be generated among the different seed runs, giving rise to a collection of centroid sites that differ in length, and making it difficult to visualize the TFBSs in a more traditional probability matrix representation.
[ "12" ]
Furthermore, different fragmentation models can be generated among the different seed runs, giving rise to a collection of centroid sites that differ in length, and making it difficult to visualize the TFBSs in a more traditional probability matrix representation.
true
true
true
true
true
1,446
17
INTRODUCTION
0
null
null
17,483,517
null
To address these two difficulties, the selection of the ‘Align Centroid Model’ option causes the Gibbs Centroid Sampler to use the Gibbs Recursive Sampler to align the collection of centroid sites.
null
197
9,075
0
false
null
null
To address these two difficulties, the selection of the ‘Align Centroid Model’ option causes the Gibbs Centroid Sampler to use the Gibbs Recursive Sampler to align the collection of centroid sites.
true
true
true
true
true
1,447
17
INTRODUCTION
0
null
null
17,483,517
null
In the case of multiple models, this process will separate the sites into related groups, and thus aid identification of the different site types.
null
146
9,076
0
false
null
null
In the case of multiple models, this process will separate the sites into related groups, and thus aid identification of the different site types.
true
true
true
true
true
1,447
17
INTRODUCTION
0
null
null
17,483,517
null
This process can also give the user insight into which positions in the models are highly conserved.
null
100
9,077
0
false
null
null
This process can also give the user insight into which positions in the models are highly conserved.
true
true
true
true
true
1,447
17
INTRODUCTION
0
null
null
17,483,517
null
It is important to note that the resulting alignment is neither a MAP alignment nor a centroid alignment of the complete set of data sequences.
null
143
9,078
0
false
null
null
It is important to note that the resulting alignment is neither a MAP alignment nor a centroid alignment of the complete set of data sequences.
true
true
true
true
true
1,447
17
INTRODUCTION
0
null
null
17,483,517
null
It is provided only to lend additional insight into the centroid solution.
null
74
9,079
0
false
null
null
It is provided only to lend additional insight into the centroid solution.
true
true
true
true
true
1,447
18
INTRODUCTION
1
3
[ "B3", "B11" ]
17,483,517
pmid-12824370|pmid-2184437
Program output is returned via e-mail.
[ "3", "11" ]
38
9,080
0
false
Program output is returned via e-mail.
[]
Program output is returned via e-mail.
true
true
true
true
true
1,448
18
INTRODUCTION
1
3
[ "B3", "B11" ]
17,483,517
pmid-12824370|pmid-2184437
The initial portion of the Gibbs Centroid Sampler output is identical to that of the other versions of the sampler, simply providing a list of the options used for the current run, followed by a list of the FASTA headings for the input sequences (see (3) for an example).
[ "3", "11" ]
271
9,081
0
false
The initial portion of the Gibbs Centroid Sampler output is identical to that of the other versions of the sampler, simply providing a list of the options used for the current run, followed by a list of the FASTA headings for the input sequences for an example).
[ "see (3" ]
The initial portion of the Gibbs Centroid Sampler output is identical to that of the other versions of the sampler, simply providing a list of the options used for the current run, followed by a list of the FASTA headings for the input sequences for an example).
true
true
true
true
true
1,448
18
INTRODUCTION
1
3
[ "B3", "B11" ]
17,483,517
pmid-12824370|pmid-2184437
Following these is the list of the sites making up the centroid model.
[ "3", "11" ]
70
9,082
0
false
Following these is the list of the sites making up the centroid model.
[]
Following these is the list of the sites making up the centroid model.
true
true
true
true
true
1,448
18
INTRODUCTION
1
11
[ "B3", "B11" ]
17,483,517
pmid-12824370|pmid-2184437
Figure 3 shows the results for a set of 18 Escherichia coli sequences; these sequences are well studied, known to contain binding sites for the cyclic AMP receptor protein (Crp) (11), and are provided as a test dataset when the Gibbs Sampler software is downloaded.
[ "3", "11" ]
265
9,083
1
false
Figure 3 shows the results for a set of 18 Escherichia coli sequences; these sequences are well studied, known to contain binding sites for the cyclic AMP receptor protein (Crp), and are provided as a test dataset when the Gibbs Sampler software is downloaded.
[ "11" ]
Figure 3 shows the results for a set of 18 Escherichia coli sequences; these sequences are well studied, known to contain binding sites for the cyclic AMP receptor protein (Crp), and are provided as a test dataset when the Gibbs Sampler software is downloaded.
true
true
true
true
true
1,448
18
INTRODUCTION
1
3
[ "B3", "B11" ]
17,483,517
pmid-12824370|pmid-2184437
The results in Figure 3 were generated using the centroid sampler with a motif width of 16, a palindromic motif model requirement, a maximum number of sites per sequence of two, heterogeneous background composition, the default number of restarts (20 seeds), the default burn-in (2000 iterations) and the default centroi...
[ "3", "11" ]
357
9,084
0
false
The results in Figure 3 were generated using the centroid sampler with a motif width of 16, a palindromic motif model requirement, a maximum number of sites per sequence of two, heterogeneous background composition, the default number of restarts (20 seeds), the default burn-in (2000 iterations) and the default centroi...
[]
The results in Figure 3 were generated using the centroid sampler with a motif width of 16, a palindromic motif model requirement, a maximum number of sites per sequence of two, heterogeneous background composition, the default number of restarts (20 seeds), the default burn-in (2000 iterations) and the default centroi...
true
true
true
true
true
1,448
18
INTRODUCTION
1
3
[ "B3", "B11" ]
17,483,517
pmid-12824370|pmid-2184437
The motif models were allowed to fragment to a width of 24 bases.
[ "3", "11" ]
65
9,085
0
false
The motif models were allowed to fragment to a width of 24 bases.
[]
The motif models were allowed to fragment to a width of 24 bases.
true
true
true
true
true
1,448
18
INTRODUCTION
1
3
[ "B3", "B11" ]
17,483,517
pmid-12824370|pmid-2184437
Figure 3.Output from the Gibbs Centroid Sampler.
[ "3", "11" ]
48
9,086
0
false
Figure 3.Output from the Gibbs Centroid Sampler.
[]
Figure 3.Output from the Gibbs Centroid Sampler.
true
true
true
true
true
1,448
19
INTRODUCTION
0
null
null
17,483,517
null
Output from the Gibbs Centroid Sampler.
null
39
9,087
0
false
null
null
Output from the Gibbs Centroid Sampler.
true
true
true
true
true
1,449
20
INTRODUCTION
1
11
[ "B11", "B2" ]
17,483,517
pmid-2184437|NA
At the top of Figure 3 is the set of sites making up the centroid; the centroid sites are listed in upper case, and flanking positions are in lower case.
[ "11", "2" ]
153
9,088
0
false
At the top of Figure 3 is the set of sites making up the centroid; the centroid sites are listed in upper case, and flanking positions are in lower case.
[]
At the top of Figure 3 is the set of sites making up the centroid; the centroid sites are listed in upper case, and flanking positions are in lower case.
true
true
true
true
true
1,450
20
INTRODUCTION
1
11
[ "B11", "B2" ]
17,483,517
pmid-2184437|NA
The sites correspond well with the DNaseI footprinted sites for these sequences (11).
[ "11", "2" ]
85
9,089
1
false
The sites correspond well with the DNaseI footprinted sites for these sequences.
[ "11" ]
The sites correspond well with the DNaseI footprinted sites for these sequences.
true
true
true
true
true
1,450
20
INTRODUCTION
1
11
[ "B11", "B2" ]
17,483,517
pmid-2184437|NA
The variation in the length of the sites is a result of different fragmentation models generated during the sampling periods (mentioned above).
[ "11", "2" ]
143
9,090
0
false
The variation in the length of the sites is a result of different fragmentation models generated during the sampling periods (mentioned above).
[]
The variation in the length of the sites is a result of different fragmentation models generated during the sampling periods (mentioned above).
true
true
true
true
true
1,450
20
INTRODUCTION
1
11
[ "B11", "B2" ]
17,483,517
pmid-2184437|NA
The dynamic program that calculates the centroid can be found elsewhere [see the supplementary material for (2)].
[ "11", "2" ]
113
9,091
0
false
The dynamic program that calculates the centroid can be found elsewhere.
[ "see the supplementary material for (2)" ]
The dynamic program that calculates the centroid can be found elsewhere.
true
true
true
true
true
1,450
20
INTRODUCTION
1
11
[ "B11", "B2" ]
17,483,517
pmid-2184437|NA
The legend below the list of sites identifies the various columns of the output.
[ "11", "2" ]
80
9,092
0
false
The legend below the list of sites identifies the various columns of the output.
[]
The legend below the list of sites identifies the various columns of the output.
true
true
true
true
true
1,450
20
INTRODUCTION
1
11
[ "B11", "B2" ]
17,483,517
pmid-2184437|NA
The probability column shows the sampling frequencies for these sites.
[ "11", "2" ]
70
9,093
0
false
The probability column shows the sampling frequencies for these sites.
[]
The probability column shows the sampling frequencies for these sites.
true
true
true
true
true
1,450
20
INTRODUCTION
1
11
[ "B11", "B2" ]
17,483,517
pmid-2184437|NA
These sampling frequencies are an estimate of the probabilities that the cognate transcription factors bind at the predicted sites.
[ "11", "2" ]
131
9,094
0
false
These sampling frequencies are an estimate of the probabilities that the cognate transcription factors bind at the predicted sites.
[]
These sampling frequencies are an estimate of the probabilities that the cognate transcription factors bind at the predicted sites.
true
true
true
true
true
1,450
21
INTRODUCTION
1
3
[ "B3" ]
17,483,517
pmid-12824370
The second part of Figure 3 shows an alignment of the centroid sites.
[ "3" ]
69
9,095
0
false
The second part of Figure 3 shows an alignment of the centroid sites.
[]
The second part of Figure 3 shows an alignment of the centroid sites.
true
true
true
true
true
1,451
21
INTRODUCTION
1
3
[ "B3" ]
17,483,517
pmid-12824370
The program generates this alignment by taking the collection of sites in the centroid, plus their flanking sequences, and using the Gibbs Recursive Sampler to find the best alignment among this set of sites, with at most one site in each sequence.
[ "3" ]
248
9,096
0
false
The program generates this alignment by taking the collection of sites in the centroid, plus their flanking sequences, and using the Gibbs Recursive Sampler to find the best alignment among this set of sites, with at most one site in each sequence.
[]
The program generates this alignment by taking the collection of sites in the centroid, plus their flanking sequences, and using the Gibbs Recursive Sampler to find the best alignment among this set of sites, with at most one site in each sequence.
true
true
true
true
true
1,451
21
INTRODUCTION
1
3
[ "B3" ]
17,483,517
pmid-12824370
As such, this is neither a centroid nor an optimal alignment.
[ "3" ]
61
9,097
0
false
As such, this is neither a centroid nor an optimal alignment.
[]
As such, this is neither a centroid nor an optimal alignment.
true
true
true
true
true
1,451
21
INTRODUCTION
1
3
[ "B3" ]
17,483,517
pmid-12824370
It is provided simply to allow the user to identify different site types (when multiple motif models were used) and to visualize which positions are highly conserved in the centroid sites.
[ "3" ]
188
9,098
0
false
It is provided simply to allow the user to identify different site types (when multiple motif models were used) and to visualize which positions are highly conserved in the centroid sites.
[]
It is provided simply to allow the user to identify different site types (when multiple motif models were used) and to visualize which positions are highly conserved in the centroid sites.
true
true
true
true
true
1,451
21
INTRODUCTION
1
3
[ "B3" ]
17,483,517
pmid-12824370
The format of this alignment is identical to that of the Gibbs Recursive Sampler previously described in (3).
[ "3" ]
109
9,099
1
false
The format of this alignment is identical to that of the Gibbs Recursive Sampler previously described in.
[ "3" ]
The format of this alignment is identical to that of the Gibbs Recursive Sampler previously described in.
true
true
true
true
true
1,451