def _vcf_parseheader(fileobj):
pop = Population()
for line in fileobj:
if line.startswith('##'):
continue
elif line.startswith('#'):
ind_ids = line.strip().split()[9:]
inds = [Individual(pop, ind_id) for ind_id in ind_ids]
for ind in inds:
pop.register_individual(ind)
return pop, inds
else:
raise FileFormatError("No header line in VCF")
def read_vcf(filename, require_pass=False, freq_info=None, info_filters=None):
'''
Reads a VCF file and returns a Population object with the
individuals represented in the file
'''
if not info_filters:
info_filters = []
for filter in info_filters:
if not callable(filter):
raise ValueError('Filter not callable')
with open(filename) as f:
pop = Population()
last_chrom = None
genotypes = []
for i, line in enumerate(f):
if line.startswith('##'):
continue
elif line.startswith('#'):
ind_ids = line.strip().split()[9:]
inds = [Individual(pop, ind_id) for ind_id in ind_ids]
for ind in inds:
pop.register_individual(ind)
break
for i, line in enumerate(f):
record = VCFRecord(line)
if info_filters and not all(filter(record) for filter in info_filters):
continue
if require_pass and not record.filter_passed:
continue
if record.chrom != last_chrom:
if last_chrom is not None:
chromobj.finalize()
pop.add_chromosome(chromobj)
chromobj = ChromosomeTemplate(label=record.chrom)
if freq_info is not None and freq_info in record.info:
freq = record.info[freq_info]
if ',' in freq:
freq = freq.split(',')[0]
freq = float(freq)
else:
freq = 0
genorow = record.genotypes()
genotypes.append(genorow)
chromobj.add_genotype(bp=record.pos,
label=record.label,
frequency=freq)
last_chrom = record.chrom
chromobj.finalize()
pop.add_chromosome(chromobj)
for ind in inds:
# Initialize new genotypes
ind._init_genotypes(sparse=True)
# Now actually sift through markers and assign them to individuals
final_indices = []
for chromidx, chromobj in enumerate(pop.chromosomes):
indices = zip([chromidx]*chromobj.nmark(), range(chromobj.nmark()))
final_indices.extend(indices)
raw_indices = range(len(genotypes))
for raw, final in zip(raw_indices, final_indices):
chromidx, markidx = final
row = genotypes[raw]
assign_genorow(row, inds, chromidx, markidx)
# Kill the row so we don't end up with the whole dataset in memory twice
genotypes[raw] = None
return pop
