def build_distance_matrix(self, fname) :
tmp = []
fq = FastqFile(fname)
fq.open()
for seq in fq :
tmp.append((seq.id[1:], seq.sequence))
fq.close()
p = Progress("Calculating distance matrix", (len(tmp)*(len(tmp)-1)) / 2.0, False)
p.start()
dist = {}
for index, labelseq in enumerate(tmp) :
label1,seq1 = labelseq
dist[(label1,label1)] = 1.0
for label2,seq2 in tmp[index+1:] :
dist[(label1,label2)] = \
dist[(label2,label1)] = \
self.distance([seq1,seq2])
p.increment()
p.end()
return dist, [ label for label, seq in tmp ]
def create_clusters(self, keys=None, homopolymer_correction=True) :
seqcount = collections.Counter()
if keys == None :
keys = self.db
for key in keys :
seqcount[key] = self.db.get(key).duplicates
p = Progress("Clustering", len(seqcount))
p.start()
for key,freq in seqcount.most_common() :
clustered = False
seq = self.db.get(key)
for c in self.clusters :
if self.alignment_similarity(self.db.get(c[0]), seq, homopolymer_correction) >= self.similarity_threshold :
c.append(key)
clustered = True
break
if not clustered :
self.clusters.append([key])
self.clusters.sort(key=len, reverse=True)
p.increment()
p.end()
def preprocess(self) :
if len(self.options['input-files']) == 0 :
self.log.error("nothing to do")
sys.exit(1)
self.seqdb = SequenceDB(preprocessed=False)
p = Progress("Preprocessing", len(self.options['input-files']))
p.start()
samples = []
for f in self.__get_files(self.options['input-files']) :
mid = self.__mid_fastq(f)
sample = Sample(f,
self.options['outdir'],
self.seqdb,
self.__filters(mid),
chimeras=self.options['chimeras'])
sample.print_sample()
samples.append(sample)
p.increment()
p.end()
rejected_reads = sum([ sum(s.filters.counts) for s in samples ])
accepted_reads = sum([ len(s) for s in samples ])
unique_seq = sum([ len(s.seqcounts) for s in samples ])
print "processed %s reads, accepted %d (of which %d are unique)" % \
(rejected_reads + accepted_reads, accepted_reads, unique_seq)
if samples :
summary_data = self.__build_summary(samples)
summary_file = Summary(self.options['summary-file'])
summary_file.update(summary_data)
summary_file.write(self.options['summary-file'])
if self.options['verbose'] :
self.summary()
return 0
def read_nematodes(self, fastq_fname, fprimer, rprimer, diffs, length) :
tmp = []
acc2name = {}
# read in sequences
fq = FastqFile(fastq_fname)
fq.open()
for seq in fq :
if 'Nematoda' not in seq.id :
continue
seq.ungap()
seq.back_translate()
new_id = seq.id.split()[0][1:]
tmp.append((new_id, seq.sequence))
acc2name[new_id] = seq.id[seq.id.find('Nematoda'):]
fq.close()
# test sequences
p = Progress("Looking for primer sequences", len(tmp), False)
p.start()
tmp2 = []
for label,seq in tmp :
findex = IUPAC.seq_position(fprimer, seq, diffs)
if findex != -1 :
#if IUPAC.seq_position_reverse(rprimer, seq, diffs) != -1 :
shortseq = seq[findex + len(fprimer) : findex + len(fprimer) + length]
if 'N' not in shortseq :
tmp2.append((label, shortseq))
p.increment()
p.end()
return tmp2,acc2name
def assign_scores(self, dist, keys, threshold, tree) :
scores = []
p = Progress("Calculating scores", len(keys), False)
p.start()
for k1 in keys :
tmp = []
for k2 in keys :
if dist[(k1,k2)] >= threshold :
tmp.append(k2)
p.increment()
#pdscore = self.get_phylogenetic_metric(tmp, tree)
scores.append((k1, 1.0, tmp))
#scores.append((k1, pdscore, tmp))
#scores.append((k1, 1 / float(len(tmp)), tmp))
p.end()
return scores
if method == 'blastlocal' :
names[name].append(acc2tax.get(fields[1], "unknown"))
else :
try :
desc = fields[1].split('|')
if re.match(".+\.\d+", desc[3]) :
names[name].append((desc[3], fields[2]))
except IndexError :
self.log.warn("could not split line from blastn: %s" % str(fields))
continue
p = Progress("OTU naming (%s)" % method, sum([len(i) for i in names.values()]) if method == 'taxonomy' else len(names))
p.start()
# now generate labels
if method == 'blast' :
for name in names :
tmp = names[name][0]
names[name] = "%s_%s" % (self.__get_desc_wrapper(tmp[0], method, 10), tmp[1])
p.increment()
elif method == 'taxonomy' :
for name in names :
names[name] = "%s" % (self.__get_desc_common(names[name], method, p))
elif method == 'blastlocal' :
for name in names :
names[name] = "%s" % (self.__merge_taxonomy(names[name]))
p.increment()