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 bipartition(reads):
positions = reads.get_positions()
# create genotypes over your variants: all heterozygous (=1)
genotypes = canonic_index_list_to_biallelic_gt_list([1] * len(positions))
# genotype likelihoods are None
genotype_likelihoods = [None] * len(positions)
# create empty pedigree
pedigree = Pedigree(NumericSampleIds())
# add one individual to pedigree
pedigree.add_individual('individual0', genotypes, genotype_likelihoods)
# recombination cost vector, irrelevant if one using one individual
recombcost = [1] * len(positions)
# run the core phasing algorithm, creating a DP table
dp_table = PedigreeDPTable(reads,
recombcost,
pedigree,
distrust_genotypes=False)
phasing, transmission_vector = dp_table.get_super_reads()
#print('PHASING')
#print(phasing[0])
#print(phasing[0][0])
#print(phasing[0][1])
mec_score = dp_table.get_optimal_cost()
eprint("MEC Score:", mec_score)
eprint("MEC Score / readset length:",
float(mec_score) / float(readset_length))
# In case the bi-partition of reads is of interest:
partition = dp_table.get_optimal_partitioning()
#print(partition)
eprint("partition fraction:", sum(partition) / float(len(partition)))
return phasing, partition
def phase_MAV(reads, n_alleles, all_het, genos, genotypes, weights=None):
readset = string_to_readset(reads, n_alleles)
positions = readset.get_positions()
for all_heterozygous in all_het:
recombcost = [1] * len(
positions) # recombination costs 1, should not occur
pedigree = Pedigree(NumericSampleIds())
genotype_likelihoods = [
None if all_heterozygous else PhredGenotypeLikelihoods(genos)
] * len(positions)
pedigree.add_individual(
'individual0', genotypes,
genotype_likelihoods) # all genotypes heterozygous
dp_table = PedigreeDPTable(readset,
recombcost,
pedigree,
distrust_genotypes=not all_heterozygous)
superreads_list, transmission_vector = dp_table.get_super_reads()
cost = dp_table.get_optimal_cost()
return superreads_list, transmission_vector, cost