def read_beagle_markerfile(filename, label=None):
'''
Reads marker locations from a beagle format file
Arguments
-----
filename: The file to be read
label: An optional label to give the chromosome, since the BEAGLE
format does not require it
Returns: a ChromosomeTemplate object
'''
with open(filename) as f:
chrom = ChromosomeTemplate(label=label)
last_pos = -1
for line in f:
rec = BeagleMarkerRecord(line)
if rec.pos < 0:
raise FileFormatError(
'Bad position for genotype: {}'.format(rec.pos))
elif rec.pos <= last_pos:
raise FileFormatError('Makers in file out of order')
chrom.add_genotype(None, cm=None, label=rec.label, bp=rec.pos,
reference=rec.reference, alternates=rec.alternates)
last_pos = rec.pos
return chrom
def direct_to_disk(filename, pop, seed_size=500,
min_length=1, size_unit='mb',
min_density=100, maxmiss=0.25,
onlywithin=False, njobs=1):
with open(filename, 'w') as outf:
results = _perform_sgs(pop, seed_size=seed_size,
min_length=min_length, size_unit=size_unit,
min_density=min_density, maxmiss=maxmiss,
onlywithin=onlywithin, njobs=njobs)
for result in results:
for segment in result.segments:
outf.write(segment.to_germline() + '\n')
def write_sgs(data, filename):
# GERMLINE files are text files with the format:
# 0) Family ID 1
# 1) Individual ID 1
# 2) Family ID 2
# 3) Individual ID 2
# 4) Chromosome
# 5) Segment start (bp/cM)
# 6) Segment end (bp/cM)
# 7) Segment start (SNP)
# 8) Segment end (SNP)
# 9) Total SNPs in segment
# 10) Genetic length of segment
# 11) Units for genetic length (cM or MB)
# 12) Mismatching SNPs in segment
# 13) 1 if Individual 1 is homozygous in match; 0 otherwise
# 14) 1 if Individual 2 is homozygous in match; 0 otherwise
with open(filename, 'w') as o:
for segment in data.segments:
oline = []
ind1 = segment.ind1.full_label
ind2 = segment.ind2.full_label
oline.extend(ind1)
oline.extend(ind2)
chrom = [segment.chromosome.label]
physical = segment.physical_location
labs = segment.marker_labels
nmark = [segment.nmark]
psize = [segment.physical_size / 1e6] # Megabases, not basepairs
oline.extend(chrom)
oline.extend(physical)
oline.extend(labs)
oline.extend(nmark)
oline.extend(psize)
unit = ['MB']
# Extra info GERMLINE gives you like mismatch rate
misc = 'X', 'X', 'X'
oline.extend(unit)
oline.extend(misc)
oline = '\t'.join([str(x) for x in oline])
o.write(oline)
o.write('\n')
def read_kinship(filename):
'''
Reads a KinInbCoef formatted file of kinship and inbreeding coefficients
Arguments
------
filename: the filename to be read
Returns: a dictionary in the format
{frozenset({(fam, ind_a), (fam, ind_b)}): kinship/inbreeding
'''
kindict = {}
with open(filename) as f:
for line in f:
fam, ida, idb, phi = line.strip().split()
kindict[frozenset({(fam, ida), (fam, idb)})] = float(phi)
return kindict
def direct_to_disk(filename,
pop,
seed_size=500,
min_length=1,
size_unit='mb',
min_density=100,
maxmiss=0.25,
onlywithin=False,
njobs=1):
with open(filename, 'w') as outf:
results = _perform_sgs(pop,
seed_size=seed_size,
min_length=min_length,
size_unit=size_unit,
min_density=min_density,
maxmiss=maxmiss,
onlywithin=onlywithin,
njobs=njobs)
for result in results:
for segment in result.segments:
outf.write(segment.to_germline() + '\n')
def read_beagle_genotypefile(filename, pop, missingcode='0'):
'''
Reads BEAGLE formatted genotype files
Arguments
------
filename: Filename of BEAGLE genotype file
pop: the population to add these individuals to
missingcode: The value that indicates a missing genotype
Returns: Nothing
'''
with open(filename) as f:
for line in f:
rec = BeagleGenotypeRecord(line)
if rec.identifier == 'I':
inds = [Individual(pop, label) for label in rec.data[::2]]
elif rec.is_phenotype_record:
for ind, pheno_status in zip(inds, rec.data[::2]):
if rec.identifier == 'A':
pheno_status = pheno_status == '2'
else:
try:
pheno_status = float(pheno_status)
except ValueError:
pass
ind.phenotypes[rec.label] = pheno_status
else:
# We've reached the genotypes, and we're skipping out
break
f.seek(0)
gtrows = [list(grouper(BeagleGenotypeRecord(x).data, 2))
for x in f if x.startswith('M')]
genotypes = zip(*gtrows)
for ind, sequentialalleles in zip(inds, genotypes):
ind.genotypes = genotypes_from_sequential_alleles(ind.chromosomes,
sequentialalleles,
missingcode=missingcode)
def read_germline(filename):
'''
Reads a GERMLINE formatted SGS filename into an SGSAnalysis object
GERMLINE files are text files with the format:
0) Family ID 1
1) Individual ID 1
2) Family ID 2
3) Individual ID 2
4) Chromosome
5) Segment start (bp/cM)
6) Segment end (bp/cM)
7) Segment start (SNP)
8) Segment end (SNP)
9) Total SNPs in segment
10) Length of segment
11) Units for genetic length (cM or MB)
12) Mismatching SNPs in segment
13) 1 if Individual 1 is homozygous in match; 0 otherwise
14) 1 if Individual 2 is homozygous in match; 0 otherwise
This function only uses 0-6.
'''
analysis = SGSAnalysis()
with open(filename) as f:
for line in f:
rec = GermlineRecord(line)
if rec.pair not in analysis:
analysis[rec.pair] = SGS(rec.ind1, rec.ind2)
phys_loc = (rec.location if rec.bp_locations else None)
gen_loc = (rec.location if not rec.bp_locations else None)
seg = Segment(rec.ind1, rec.ind2, rec.chromosome, None, None,
physical_location=phys_loc)
analysis[rec.pair].append(seg)
return analysis
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
