BayesMotif: Sorting motif discovery using Bayes Classifiers (version 1.0)

see Manual

Citation note:If you use BayesMotif in your publications, please cite the following publication:
(1) Jianjun Hu and Fan Zhang. BayesMotif: de novo protein sorting motif discovery from impure datasets. BMC Bioinformatics 2010.11(Suppl 1):S66
Introduction: BayesMotif is a de novo identification algorithm for finding a common type of protein sorting motifs in which a highly conserved anchor is present along with a less conserved motif regions. It first screens out over-represented anchors from the positive protein seqequence dataset and then build Bayesian classifers around each highly conserved anchor. It then ranks the anchors by the classification accuracy of that anchor along with surrounding context to differentiate positive sequences from negative sequences. BayesMotif can also be used to find other type of protein motifs if they can be assumed to contain a highly conserved cores in the motif

Experiment Preparation:: Users need to prepare two sets of protein sequence fasta files and to specify the following parameters based on the prior knowledge of the problem:

  • Positive protein sequence set (fasta format protein sequence file): these sequences are assumed to contain the likely motifs of interests. While some of the sequences may not actually contain the motifs, a majority of them
  • Negative protein sequence set: these sequences are assumed to not contain the potential motifs.
  • Key parameters: anchor (regular expression) model, left and right scanning regions around anchors


Please enter your email address, where the prediction will be sent to:

PositiveSet: Paste the positive sequences in FASTA format below or upload a fasta file:
(download a positive sample file )

NegativeSet: Paste the background/negative sequences in FASTA format below or upload a fasta file: (download a negative sample file )


Discover motifs at N-Terminal C-Terminal
Apply False Positive Removal No False Positive Removal
  




Todo tasks:
  • set all parameters for the algorithm from webform.
  • output weblogo for the discovered motifs.

Last Modified: Feb 28, 2010