Does any bioc package support amino acid sequence search using a pre- defined position weight matrix (pwm)? As best I can tell, Biostrings, for all of its capabilities, does not support sequence matching using pwm's. cosmo (again, as best I can tell) does not accept a pwm as input; rather, it identifies pwm's from the input sequences. I'd be delighted to find out I am wrong on one or both counts ... or to hear of another bioc package that does sequence search with pwm's. For background: I obtain position weight matrices which predict kinase substrates from Predikin (http://predikin.biosci.uq.edu.au/ pkr/). I wish to use these matrices to search for possible kinase substrates in peptide sequences obtained from mass spectrometry. Many thanks -- - Paul

Hi Paul, I don't think bioc does (though someone can correct me). If you turn the pwm matrix 90 degrees clockwise, you get a Position Specific Scoring Matrix (like a Gribskov profile, but without gap scores), which can be matched to a string using the Viterbi algorithm. There are ways to execute the Viterbi algorithm in R, see for example: http://tolstoy.newcastle.edu.au/R/help/06/03/23821.html Regards, Paul > Does any bioc package support amino acid sequence search using a pre- > defined > position weight matrix (pwm)? > > As best I can tell, Biostrings, for all of its capabilities, does not > support > sequence matching using pwm's. cosmo (again, as best I can tell) > does not > accept a pwm as input; rather, it identifies pwm's from the input > sequences. > > I'd be delighted to find out I am wrong on one or both counts ... or > to hear > of another bioc package that does sequence search with pwm's. > > For background: I obtain position weight matrices which predict > kinase substrates from Predikin (http://predikin.biosci.uq.edu.au/ > pkr/). I wish > to use these matrices to search for possible kinase substrates in > peptide sequences > obtained from mass spectrometry. > > Many thanks -- > > - Paul > > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor > > >

Hi Paul, No there is no support for PWMs in Biostrings. To deal with ambiguities there is support for the use of extended IUPAC letters in the pattern (see the 'fixed' arg in ?matchPattern) but unfortunately this is for a pattern of nucleotides, not for amino acids. Cheers, H. Paul Shannon wrote: > Does any bioc package support amino acid sequence search using a pre- > defined > position weight matrix (pwm)? > > As best I can tell, Biostrings, for all of its capabilities, does not > support > sequence matching using pwm's. cosmo (again, as best I can tell) > does not > accept a pwm as input; rather, it identifies pwm's from the input > sequences. > > I'd be delighted to find out I am wrong on one or both counts ... or > to hear > of another bioc package that does sequence search with pwm's. > > For background: I obtain position weight matrices which predict > kinase substrates from Predikin (http://predikin.biosci.uq.edu.au/ > pkr/). I wish > to use these matrices to search for possible kinase substrates in > peptide sequences > obtained from mass spectrometry. > > Many thanks -- > > - Paul > > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor