def genotype_pedigree(
numeric_sample_ids,
reads,
recombcost,
pedigree,
expected_genotypes,
weights=None,
expected=None,
scaling=10,
positions=None,
):
rs = string_to_readset_pedigree(s=reads, w=weights, scaling_quality=scaling)
dp_forward_backward = GenotypeDPTable(numeric_sample_ids, rs, recombcost, pedigree, positions)
# for each position compare the likeliest genotype to the expected ones
print("expected genotypes: ", expected_genotypes)
positions = rs.get_positions()
for pos in range(len(positions)):
for individual in range(len(pedigree)):
likelihoods = dp_forward_backward.get_genotype_likelihoods(
"individual" + str(individual), pos
)
# if expected likelihoods given, compare
if expected is not None:
print(
"likelihoods: ",
likelihoods,
" expected likelihoods: ",
expected[individual][pos],
)
assert likelihoods == PhredGenotypeLikelihoods(expected[individual][pos])
# find the likeliest genotype
max_val = -1
max_geno = Genotype([])
for genotype in likelihoods.genotypes():
assert not math.isnan(likelihoods[genotype])
if likelihoods[genotype] > max_val:
max_val = likelihoods[genotype]
max_geno = genotype
# compare it to the expected genotype
print(
"pos.: " + str(pos) + " individual " + str(individual) + ": ",
likelihoods,
" expected genotype: ",
expected_genotypes[individual][pos],
)
assert max_geno == expected_genotypes[individual][pos]
print("\n")
def phase_pedigree(reads,
recombcost,
pedigree,
distrust_genotypes=False,
positions=None):
rs = string_to_readset_pedigree(reads)
dp_table = PedigreeDPTable(rs, recombcost, pedigree, distrust_genotypes,
positions)
superreads_list, transmission_vector = dp_table.get_super_reads()
cost = dp_table.get_optimal_cost()
for superreads in superreads_list:
for sr in superreads:
print(sr)
print("Cost:", dp_table.get_optimal_cost())
print("Transmission vector:", transmission_vector)
print("Partition:", dp_table.get_optimal_partitioning())
return superreads_list, transmission_vector, cost
def genotype_pedigree(numeric_sample_ids,
reads,
recombcost,
pedigree,
expected_genotypes,
weights=None,
expected=None,
scaling=10,
positions=None):
rs = string_to_readset_pedigree(s=reads,
w=weights,
scaling_quality=scaling)
dp_forward_backward = GenotypeDPTable(numeric_sample_ids, rs, recombcost,
pedigree, positions)
# for each position compare the likeliest genotype to the expected ones
print('expected genotypes: ', expected_genotypes)
positions = rs.get_positions()
for pos in range(len(positions)):
for individual in range(len(pedigree)):
likelihoods = dp_forward_backward.get_genotype_likelihoods(
'individual' + str(individual), pos)
# if expected likelihoods given, compare
if expected is not None:
print('likelihoods: ', likelihoods, ' expected likelihoods: ',
expected[individual][pos])
assert (likelihoods == expected[individual][pos])
# find the likeliest genotype
max_val = -1
max_index = -1
for i in range(len(likelihoods)):
assert (not math.isnan(likelihoods[i]))
if likelihoods[i] > max_val:
max_val = likelihoods[i]
max_index = i
# compare it to the expected genotype
print(
'pos.: ' + str(pos) + ' individual ' + str(individual) + ': ',
likelihoods, ' expected genotype: ',
expected_genotypes[individual][pos])
assert (max_index == expected_genotypes[individual][pos])
print("\n")