def run(self):
self.getAlignmentDict()
valueDict = {}
for aId, aln in self.alignmentDict.iteritems():
valueDict[aId] = format_ratio(aln.matches + aln.repMatches, aln.matches + aln.repMatches + aln.misMatches +
aln.qNumInsert)
self.dumpValueDict(valueDict)
def bam_is_paired(path, num_reads=100000, paired_cutoff=0.75):
"""
Infers the paired-ness of a bam file.
"""
sam = pysam.Samfile(path)
count = 0
for i, rec in enumerate(sam):
if rec.is_paired:
count += 1
if i == num_reads:
break
if format_ratio(count, num_reads) > 0.75:
return True
elif format_ratio(count, num_reads) < 1 - paired_cutoff:
return False
else:
raise RuntimeError("Unable to infer pairing from bamfile {}".format(path))
def run(self):
self.get_fasta()
results_dict = {}
for aln_id, t in self.transcript_iterator():
cds = t.get_cds(self.seq_dict)
v = 100 * format_ratio(cds.count("N"), len(cds))
results_dict[aln_id] = v
self.dump_attribute_results_to_disk(results_dict)
def has_only_short(bins,
ids_included,
ref_interval,
tgt_intervals,
percentage_of_ref=60.0):
"""
Are all of the consensus transcripts we found for this gene too short?
"""
source_size = len(ref_interval)
tgt_sizes = [
len(tgt_intervals[x]) for x in zip(*bins.itervalues())[0]
if x in ids_included
]
return all([
100 * format_ratio(tgt_size, source_size) < percentage_of_ref
for tgt_size in tgt_sizes
])
def coverage(self):
return 100 * format_ratio(self.matches + self.mismatches + self.repmatches, self.q_size)
def percent_n(self):
return 100 * format_ratio(self.n_count, self.q_size)
def target_coverage(self):
return 100 * format_ratio(self.matches + self.mismatches + self.repmatches, self.t_size)
def identity(self):
return 100 * format_ratio(self.matches + self.repmatches,
self.matches + self.repmatches + self.mismatches + self.q_num_insert)
def identity(p_list):
m = sum(x.matches for x in p_list)
mi = sum(x.misMatches for x in p_list)
rep = sum(x.repMatches for x in p_list)
ins = sum(x.qNumInsert for x in p_list)
return format_ratio(m + rep, m + rep + mi + ins)
def coverage(p_list):
m = sum(x.matches for x in p_list)
mi = sum(x.misMatches for x in p_list)
rep = sum(x.repMatches for x in p_list)
return format_ratio(m + mi + rep, p_list[0].qSize)
