def main() :
if (len(argv) < 2) or (argv[1] in ('-h', '--help', 'help')) :
usage()
return 1
command = argv[1]
if command not in get_commands() :
print >> stderr, "Error: unknown command '%s'" % command
usage()
return 1
if command == 'test' :
test_system(output=True)
return 0
options = parse_args(command, argv[2:])
log = setup_logging(options['verbose'])
check_options(command, options, log)
test_system(command, options, exit_on_failure=True)
System.tempdir(options['outdir']) # some objects need this set
wf = WorkFlow(options)
if command == 'preprocess' :
return wf.preprocess()
elif command == 'summary' :
return wf.summary()
elif command == 'cluster' :
return wf.cluster()
elif command == 'label' :
return wf.label()
elif command == 'showcounts' :
return wf.showcounts()
elif command == 'showlabels' :
return wf.showlabels()
elif command == 'phylogeny' :
return wf.phylogeny()
elif command == 'heatmap' :
return wf.heatmap()
elif command == 'wasabi' :
return wf.wasabi()
else :
print >> stderr, "'%s' appears to be partially implemented!" % command
return 1
def alignment_similarity(self, seq1, seq2, homopolymer_correction) :
# write out
f = open(System.tempfilename(ext='cluster'), 'w')
print >> f, seq1.fasta()
print >> f, seq2.fasta()
f.close()
# align
aligned = []
if homopolymer_correction :
fq = Pagan().get_454_alignment(f.name)
else :
fq = Pagan().get_alignment(f.name)
fq.open()
for seq in fq :
if seq.id == ">consensus" :
continue
aligned.append(seq.sequence)
fq.close()
# delete tmp files
os.remove(f.name)
os.remove(fq.get_filename())
# if things are really dissimilar they do not align
# so just give up here for this cluster
if len(aligned) != 2 :
return 0.0
return self.distance2(aligned, homopolymer_correction)
def check_options(command, options, log) :
system = System()
apply_prefix(options, command)
# for i in options :
# print i, options[i]
if (command != 'preprocess') and (not system.check_directory(options['outdir'])) :
exit(1)
if command == 'preprocess' :
if not system.check_directory(options['outdir'], create=True) :
exit(1)
if not system.check_files(options['input-files']) :
exit(1)
if options['denoise'] :
if not options['forwardprimer'] :
log.error("for denoising you must specify the forward primer!")
exit(1)
elif command == 'cluster' :
#if options['metadata'] is None :
# print >> stderr, "Error: you must specify a metadata file"
# exit(1)
if (options['metadata'] is not None) and (not system.check_file(options['metadata'])) :
exit(1)
for i in ('duplicate-threshold', 'total-duplicate-threshold', 'sample-threshold') :
if options[i] <= 0 :
log.error("%s must be > 0 (read %d)" % (i, options[i]))
exit(1)
# i think pagan's sensitivity limit is ~80%
if options['otu-similarity'] < 0.8 or options['otu-similarity'] > 1.0 :
log.error("similarity must be between 0.8 and 1.0 (read %.2f)" % options['otu-similarity'])
exit(1)
if options['labels'] :
if options['labels'] == 'blastlocal' :
if not options['labels_db'] :
log.error("'blastlocal' requires you specify a FASTA file to use as a blast database using the --dbfile option")
exit(1)
if not system.check_file(options['labels_db']) :
log.error("could not find %s" % options['labels_db'])
exit(1)
elif command == 'summary' :
if not system.check_file(options['summary-file']) :
log.error("could not find %s, did you run the 'preprocess' command yet?" % options['summary-file'])
exit(1)
elif command == 'label' :
fasta_check = system.check_file(options['cluster-fasta'])
biom_check = system.check_file(options['cluster-biom'])
def filename_or() :
if fasta_check ^ biom_check :
return "'%s'" % (options['cluster-fasta'] if not fasta_check else options['cluster-biom'])
else :
return "'%s' or '%s'" % (options['cluster-fasta'], options['cluster-biom'])
if not fasta_check or not biom_check :
log.error("could not find %s, did you run the 'cluster' command yet?" % filename_or())
exit(1)
if not options['labels'] :
log.error("you must specify a labelling method")
exit(1)
if options['labels'] == 'blastlocal' :
if not options['labels_db'] :
log.error("'blastlocal' requires you specify a FASTA file to use as a blast database using the --dbfile option")
exit(1)
if not system.check_file(options['labels_db']) :
log.error("could not find %s" % options['labels_db'])
exit(1)
elif command == 'showcounts' or command == 'showlabels' :
if not system.check_file(options['cluster-biom']) :
log.error("could not find '%s', did you run the 'cluster' command yet?" % options['cluster-biom'])
exit(1)
elif command == 'phylogeny' :
if not system.check_file(options['cluster-fasta']) :
exit(1)
if not options['denovo'] :
if not options['silva-fasta'] or not options['silva-tree'] :
log.error("you must either specify the location of the reference alignment and phylogeny or else use the --denovo option")
exit(1)
else :
if not system.check_files([options['silva-fasta'], options['silva-tree']]) :
exit(1)
elif command == 'heatmap' :
if not system.check_files([options['cluster-biom']]) :
exit(1)
elif command == 'wasabi' :
if not options['wasabi-user'] :
log.error("you must specify your wasabi username!")
exit(1)
if not system.check_files([options['phylogeny-xml']]) :
exit(1)
def test_system(command=None, options=None, exit_on_failure=False, output=False) :
binaries = {
'preprocess' : {
'*' : ['sff2fastq'],
'chimeras' : ['uchime'],
'denoise' : ['PyroDist', 'FCluster', 'PyroNoise'],
},
'cluster' : {
'*' : ['pagan'],
'labels' : ['blastn', 'makeblastdb']
},
'label' : {
'*' : ['blastn', 'makeblastdb']
},
'summary' : {},
'showcounts' : {},
'showlabels' : {},
'phylogeny' : {
'*' : ['pagan', 'exonerate', 'bppphysamp'],
'denovo' : ['raxml']
},
'heatmap' : {},
'wasabi' : {},
'test' : {}
}
def print_out(s) :
if output :
print s
else :
if 'not found' in s :
print >> stderr, s.lstrip()
fail = False
for b in [command] if command else binaries :
if binaries[b] :
print_out("checking system for %s command dependancies :" % bold(b))
for o in binaries[b] :
if o == '*' or not options or options[o] :
for p in binaries[b][o] :
installed = System.is_installed(p)
print_out(" %s %s%s" % (p, "" if o == "*" else "(needed for --%s) " % o,
bold_green("found.") if installed else bold_red("not found!")))
if not installed :
fail = True
print_out("")
# for some reason using pip to install cairo always fails
# so check for it here instead
# print "checking for python modules :"
# try :
# import cairo
# print " pycairo (needed by 'heatmap' command) " + bold_green("found.")
# except ImportError :
# print " pycairo (needed by 'heatmap' command) " + bold_red("not found!")
# fail = True
# print ""
if exit_on_failure and fail :
exit(1)
