A short script for running SignalP over a directory of sequence files.

signalp.py

import os
from optparse import OptionParser

def main(file, path):
    """runs SignalP for every sequence file in a directory"""
    for filename in os.listdir(file):
        #other checks other than filer suffix would be more sophisticated
        if filename.endswith('.fa') or filename.endswith('.fasta'):
            sig = signalp(path, 'gram+', os.path.join(file, filename))
            length = len(sig)
            #this prints the prediction, and the signal peptide probability
            print sig[length-4], sig[length-3]



def signalp(path, organism_type, sequence_file):
    """method to run signalp and capture the output, which it returns"""
    p = os.popen(os.path.join(path, 'signalp')+' -t '+organism_type+' '+sequence_file)
    output = p.readlines()
    return output

if __name__ == '__main__':
    parser = OptionParser(usage="Usage: %prog [options] sequence_directory",
            version="%prog 0.1")
    parser.add_option("-s", "--signalp", dest="signalp_path",
            help="Path to signalp binary, defaults to /usr/local/bin", metavar="INT")
    (options, args) = parser.parse_args()
    if options.signalp_path:
        sigp = options.signalp_path
    else:
        sigp = '/usr/local/bin'
    if len(args) != 1:
        print 'Usage error, no infile defined'
        parser.print_help()
    else:
        main(args[0], sigp)

A short script for running SignalP over a directory of sequence files.