Marleen Claeys, Valerie Storms, Hong Sun, Tom Michoel and Kathleen MarchalMarleen Claeys (mclaeys@qatar.net.qa)Kathleen Marchal (kathleen.marchal@biw.kuleuven.be)http://bioinformatics.psb.ugent.be/webtools/MotifSuite/usemotifsampler.phpM. Claeys; V. Storms; H. Sun; T. Michoel; K. Marchal. (2012) "MotifSuite: workflow for probabilistic motif detection and assessment", <i>Bioinformatics</i>, <b>28</b>(14):1931-1932Thijs G., Marchal K., Lescot M., Rombauts S., De Moor B., Rouze P., Moreau Y. (2002) "A Gibbs Sampling method to detect over-represented motifs in upstream regions of coexpressed genes", <i>Journal of Computational Biology (special issue Recomb'2001)</i>, <b>9</b>(2):447-464Thijs G., Moreau Y., De Smet F., Mathys J., Lescot M., Rombauts S., Rouze P., De Moor B., and Marchal K. (2002) "INCLUSive: INtegrated Clustering, Upstream sequence retrieval and motif Sampling", <i>Bioinformatics</i>, <b>18</b>(2):331-2
MotifSampler is a probabilistic de novo motif detection tool searching for putative motifs in DNA sequences upstream of coregulated genes from one species. Detection is done by means of a stochastic optimization strategy (a Gibbs sampling approach) that searches for all possible sets of short segments that are overrepresented in the sequence dataset compared to the surrounding nucleotides (also called the non-functional background).
<br><br>
<a href="http://bioinformatics.psb.ugent.be/webtools/MotifSuite/pPriorDistrs.php">See here for a description of the <b>Priors distribution</b> parameter</a>
%PROGRAM {Sequence} {Background} {Width} {Number} {Max occurrences} {Prior distribution} {Use positional priors} {Positional priors track} {Overlap} {Strand} {Repeat} {Sampling} {Result} {Motifs}RelativeA background model1308Width of motifs to search for1301The number of different motifs to search for01002Maximum number of instances to search for in any sequence (0 = unset).<html>Prior distribution on the number of instances per sequence to search for (-p).<br>See the MotifSampler web site for a description of this parameter</html>0.9_0.25<html>Parameters for the prior distribution.<br>The value of this parameter should be set according to<br>the chosen distribution type as described on the<br>MotifSampler web site.</html>No<html>Select steps in which to apply Position Specific Prior correction:<br>updating step, sampling step, both of these or not at all.<br>If the option to apply PSP correction is selected (value is not "No"),<br>the track to use as positional priors must also be specified.</html>RelativefalseA positional priors track which will be used if 'Use positional priors' is not 'No'Search single strand or both strands0301Sets allowed overlap between different motifs (when searching for multiple motifs)01000100Number of times one MotifSampler run should be repeated with the same parameter settings on the same input sequences dataset. Default=100<html>Default (1) will sample the number of instances that will be allocated per sequence<br>from an (internally computed) distribution on the number of instances per sequence.<br>Sampling provides a more randomized search in the solution space of all possible motifs<br>compared to mode (0) that computes the weighted average of the distribution</html>Relative1Relative{SEQUENCENAME}\t{FEATURE}\tmisc_feature\t{START}\t{END}\t{SCORE}\t{STRAND}\t.\tid "{TYPE}"; site "{SITE}";