by Jens Keilwagen, Jan Baumbach, Thomas A. Kohl and Ivo Grosse.
Valuable binding site annotation data are stored in databases. However, several types of errors can, and do, occur in the process of manually incorporating annotation data from scientific literature into these databases. Here, we introduce MotifAdjuster, a software that helps to detect these errors, and we demonstrate its efficacy on public data sets.
The paper MotifAdjuster: a tool for computational reassessment of transcription factor binding site annotations has been published in Genome Biology.
MotifAdjuster can be downloaded here.
If you have unzipped the archive, you can start the MotifAdjuster by invoking
java -cp ./:./jstacs-1.2.2.jar:./numericalMethods.jar MotifAdjuster <file> <ignoreChar> <length> <fgOrder> <fgEss> <bothStrands> <output> <sigma> <p(no motif)>
In Windows, you have to use ";" instead of ":" in the class path.
The arguments have the following meaning
|file||the location of the data set||String|
|ignoreChar||char for comment lines (e.g. for a FastA-file '>')||char|
|length||the motif length||int|
|fgOrder||the order of the inhomogeneous Markov model that is uses for the motif; 0 yields in a PWM||byte|
|ess||the equivalent sample size that is used for the mixture model||double >= 0|
|bothStrands||use both strands||boolean|
|output||output of the EM||boolean|
|sigma||the sigma of the truncated discrete Gaussian distribution||double>0|
|p(no motif)||the probability for finding no motif||0<=double<1|
The authors of SeSiMCMC informed us that the performance of SeSiMCMC could be improved by using the latest version of SeSiMCMC (version 4.31 from 2009-04-23) and by using the advanced option Background model is common (static) set to the value TRUE. Hence, we provide an updated version of additional file 1.