def test_labelledallele_delabeler():
ngenos = 10 # Number of genotypes per chromosome
if ngenos % 2 == 1:
raise ValueError('Even number of genotypes needed')
p = Population()
c = ChromosomeTemplate()
for i in range(ngenos):
c.add_genotype()
p.add_chromosome(c)
a = Individual(p, 1)
a._init_genotypes(blankchroms=False)
a.genotypes[0][0] = Alleles([1] * ngenos)
a.genotypes[0][1] = Alleles([2] * ngenos)
b = Individual(p, 2)
b._init_genotypes(blankchroms=False)
b.genotypes[0][0] = Alleles([3] * ngenos)
b.genotypes[0][1] = Alleles([4] * ngenos)
chromatid_spans = [
InheritanceSpan(a, 0, 0, 0, ngenos // 2),
InheritanceSpan(b, 0, 1, ngenos // 2, ngenos)
]
chromatid = LabelledAlleles(spans=chromatid_spans, chromobj=c)
expected_value = [1] * (ngenos // 2) + [4] * (ngenos // 2)
expected_value = Alleles(expected_value)
actual_value = chromatid.delabel()
assert all(actual_value == expected_value)
def test_labelledallele_delabeler():
ngenos = 10 # Number of genotypes per chromosome
if ngenos % 2 == 1:
raise ValueError('Even number of genotypes needed')
p = Population()
c = ChromosomeTemplate()
for i in range(ngenos):
c.add_genotype()
p.add_chromosome(c)
a = Individual(p, 1)
a._init_genotypes(blankchroms=False)
a.genotypes[0][0] = Alleles([1]*ngenos)
a.genotypes[0][1] = Alleles([2]*ngenos)
b = Individual(p, 2)
b._init_genotypes(blankchroms=False)
b.genotypes[0][0] = Alleles([3] * ngenos)
b.genotypes[0][1] = Alleles([4] * ngenos)
chromatid_spans = [InheritanceSpan(a, 0, 0, 0, ngenos//2),
InheritanceSpan(b, 0, 1, ngenos//2, ngenos)]
chromatid = LabelledAlleles(spans=chromatid_spans, chromobj=c)
expected_value = [1]*(ngenos//2) + [4] * (ngenos//2)
expected_value = Alleles(expected_value)
actual_value = chromatid.delabel()
assert all(actual_value == expected_value)
def __init__(self, label=None):
"""
Create a pedigree.
:param label: pedigree label
"""
Population.__init__(self)
self.label = label
self.kinmat = {}
self.fratmat = {}
def test_labelledalleles():
IS = InheritanceSpan
ngenos = 50
p = Population()
c = ChromosomeTemplate()
for i in range(ngenos):
c.add_genotype()
p.add_chromosome(c)
a = Individual(p, 1)
actual = LabelledAlleles.founder_chromosome(a, 0, 0, chromobj=c)
expected = LabelledAlleles(spans=[IS(a, 0, 0, 0, ngenos)], chromobj=c)
assert actual == expected
def test_labelledalleles():
IS = InheritanceSpan
ngenos = 50
p = Population()
c = ChromosomeTemplate()
for i in range(ngenos):
c.add_genotype()
p.add_chromosome(c)
a = Individual(p, 1)
actual = LabelledAlleles.founder_chromosome(a, 0, 0, chromobj=c)
expected = LabelledAlleles(spans=[IS(a, 0, 0, 0, ngenos)], chromobj=c)
assert actual == expected
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_beagle(genofile, markerfile):
'''
Reads BEAGLE formatted genotype data
:param genofile: Filename containing genotype information for individuals
:param markerfile: Filename containing marker location and allele
information corresponding to genofile
:type genofile: string
:type markerfile: string
:rtype: Population
'''
pop = Population()
chrom = read_beagle_markerfile(markerfile)
chrom.finalize()
pop.chromosomes.add_chromosome(chrom)
read_beagle_genotypefile(genofile, pop)
return pop
def read_ped(filename,
population=None,
delimiter=None,
affected_labels=None,
population_handler=None,
data_handler=None,
connect_inds=True,
onlyinds=None):
"""
Reads a plink format pedigree file, ie:
::
familyid indid father mother sex whatever whatever whatever
into a pydigree pedigree object, with optional population to
assign to pedigree members. If you don't provide a population
you can't simulate genotypes!
:param filename: The file to be read
:param population: The population to assign individuals to
:param delimiter: a string defining the field separator,
default: any whitespace
:param affected_labels: The labels that determine affection status.
:param population_handler: a function to set up the population
:param data_handler: a function to turn the
data into useful individual information
:param connect_inds: build references between individuals. Requires all
individuals be present in the file
:param onlyinds: only include data for specified individuals
:type filename: string
:type population: Population
:type delimiter: string
:type affected_labels: dict (str -> value)
:type data_handler: callable
:type connect_inds: bool
:type onlyinds: iterable
:returns: individuals contained in the pedigree file
:rtype: PedigreeCollection
"""
if not affected_labels:
affected_labels = {
'1': 0,
'2': 1,
'A': 1,
'U': 0,
'X': None,
'-9': None
}
if not isinstance(data_handler, Callable):
data_handler = lambda *x: None
if not isinstance(population_handler, Callable):
population_handler = lambda *x: None
population = Population() if population is None else population
p = Pedigree()
population_handler(p)
# Step 1: Read the data and create the individuals
with smartopen(filename) as f:
# Parse the lines in the file
for line in f:
rec = PEDRecord(line, delimiter)
if onlyinds and (rec.ind_id not in onlyinds):
continue
ind = rec.create_individual(population)
ind.pedigree = p
ind.phenotypes['affected'] = affected_labels.get(rec.aff, None)
p[ind.label] = ind
if rec.data:
data_handler(p[ind.label], rec.data)
# Step 2: Create the between-individual relationships
# Fix the individual-level data: individuals currently only have parent-ids
# in their parent fields and not references to actual individuals
if connect_inds:
connect_individuals(p)
# Step 3: Separate the individuals into pedigrees
pc = sort_pedigrees(p.individuals, population_handler)
return pc
def __init__(self, label=None):
Population.__init__(self)
self.label = label
self.kinmat = {}
self.fratmat = {}
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
