def summarize_population_sizes(dict_of_sizes):
results = {}
for pop, sizes, in dict_of_sizes.iteritems():
results[pop + '.sample_count.mean'] = fstats._mean_(sizes)
results[pop + '.sample_count.stdev'] = fstats._stdev_(sizes)
return results
def summarize_population_sizes(dict_of_sizes):
results = {}
for pop, sizes, in dict_of_sizes.iteritems():
results[pop + '.sample_count.mean'] = fstats._mean_(sizes)
results[pop + '.sample_count.stdev'] = fstats._stdev_(sizes)
return results
def calc_slice_stats(data):
"""Main function for caculating statistics.
Make it easy to add more statistics.
"""
tabix_slice, chrm, start, stop, populations, header, min_samples = data
#progress_meter(starting_time, chrm, stop, bp_processed, total_bp_in_dataset)
if tabix_slice == None or len(tabix_slice) == 0: # skip empty alignments
return None
else:
# Create lists to store values for final printing.
output_line = [chrm, start, stop]
total_depth = []
snp_wise_f_statistics = []
population_sizes = defaultdict(list)
Hs_est_dict = {}
Ht_est_dict = {}
snp_count = 0
for count, line in enumerate(tabix_slice):
vcf_line_dict = parse_vcf_line(line, header)
# CREATE FILTERS HERE:
if vcf_line_dict["FILTER"] != 'PASS':
continue
# COUNT SAMPLES IN EACH POPULATION
for pop, size in get_population_sizes(vcf_line_dict,
populations).iteritems():
population_sizes[pop].append(size)
# CALCULATE SNPWISE F-STATISTICS
allele_counts = calc_allele_counts(populations, vcf_line_dict)
f_statistics = calc_fstats(allele_counts)
# UPDATE Hs AND Ht DICTIIONARIES
Hs_est_dict, Ht_est_dict = update_Hs_and_Ht_dicts(
f_statistics, Hs_est_dict, Ht_est_dict)
f_statistics['LOCATION'] = (chrm, start, stop)
snp_wise_f_statistics.append(f_statistics)
total_depth.append(int(vcf_line_dict['INFO']["DP"]))
snp_count = count
# SUMMARIZE POPULATION WIDE STATISTICS
pop_size_statistics = summarize_population_sizes(population_sizes)
multilocus_f_statistics = calc_multilocus_f_statistics(
Hs_est_dict, Ht_est_dict)
# SKIP SAMPLES WITH TOO MANY NANs
if len(multilocus_f_statistics.values()) == 0:
return None
elif multilocus_f_statistics.values()[0] is None:
return None
else:
# UPDATE OUTPUT LINE WITH DEPTH INFO
output_line += [
snp_count,
fstats._mean_(total_depth),
fstats._stdev_(total_depth)
]
return ([chrm, start, stop, snp_count, fstats._mean_(total_depth), fstats._stdev_(total_depth)], \
pop_size_statistics, multilocus_f_statistics)
def calc_slice_stats(data):
"""Main function for caculating statistics.
Make it easy to add more statistics.
"""
tabix_slice, chrm, start, stop, populations, header, min_samples = data
#progress_meter(starting_time, chrm, stop, bp_processed, total_bp_in_dataset)
if tabix_slice == None or len(tabix_slice) == 0: # skip empty alignments
return None
else:
# Create lists to store values for final printing.
output_line = [chrm, start, stop]
total_depth = []
snp_wise_f_statistics = []
population_sizes = defaultdict(list)
Hs_est_dict = {}
Ht_est_dict = {}
snp_count = 0
for count, line in enumerate(tabix_slice):
vcf_line_dict = parse_vcf_line(line, header)
# CREATE FILTERS HERE:
if vcf_line_dict["FILTER"] != 'PASS':
continue
# COUNT SAMPLES IN EACH POPULATION
for pop, size in get_population_sizes(vcf_line_dict, populations).iteritems():
population_sizes[pop].append(size)
# CALCULATE SNPWISE F-STATISTICS
allele_counts = calc_allele_counts(populations, vcf_line_dict)
f_statistics = calc_fstats(allele_counts)
# UPDATE Hs AND Ht DICTIIONARIES
Hs_est_dict, Ht_est_dict = update_Hs_and_Ht_dicts(f_statistics, Hs_est_dict, Ht_est_dict)
f_statistics['LOCATION'] = (chrm, start, stop)
snp_wise_f_statistics.append(f_statistics)
total_depth.append(int(vcf_line_dict['INFO']["DP"]))
snp_count = count
# SUMMARIZE POPULATION WIDE STATISTICS
pop_size_statistics = summarize_population_sizes(population_sizes)
multilocus_f_statistics = calc_multilocus_f_statistics(Hs_est_dict, Ht_est_dict)
# SKIP SAMPLES WITH TOO MANY NANs
if len(multilocus_f_statistics.values()) == 0:
return None
elif multilocus_f_statistics.values()[0] is None:
return None
else:
# UPDATE OUTPUT LINE WITH DEPTH INFO
output_line += [snp_count, fstats._mean_(total_depth), fstats._stdev_(total_depth)]
return ([chrm, start, stop, snp_count, fstats._mean_(total_depth), fstats._stdev_(total_depth)], \
pop_size_statistics, multilocus_f_statistics)
