def hom_alt_lik(self, variant_probe_coverage):
hom_alt_liks = []
# Either alt+cov or alt_covg + contam_covg
# logger.debug("hom alt %s" % variant_probe_coverage.var_name)
for expected_depth in self.expected_depths:
kmer_count_likelihood = log_lik_R_S_kmer_count(
variant_probe_coverage.alternate_kmer_count,
variant_probe_coverage.reference_kmer_count,
self.depth_to_expected_kmer_count(
expected_depth, variant_probe_coverage.reference_klen),
self.depth_to_expected_kmer_count(
expected_depth * self.error_rate / 3,
variant_probe_coverage.alternate_klen))
ngaps_likelihood=log_lik_probability_of_N_gaps(expected_depth * self.error_rate / 3,
variant_probe_coverage.reference_percent_coverage,
self.klen_to_dna_len(variant_probe_coverage.reference_klen))+\
log_lik_probability_of_N_gaps(expected_depth,
variant_probe_coverage.alternate_percent_coverage,
self.klen_to_dna_len(variant_probe_coverage.alternate_klen))
logger.debug(
"hom alt kmer_count_likelihood: %f %s" %
(kmer_count_likelihood, variant_probe_coverage.var_name))
logger.debug("hom alt ngaps_likelihood: %f %s " %
(ngaps_likelihood, variant_probe_coverage.var_name))
hom_alt_liks.append(kmer_count_likelihood + ngaps_likelihood)
# for contamination in self.contamination_depths:
# hom_alt_liks.append(
# log_lik_R_S_kmer_count(
# variant_probe_coverage.alternate_kmer_count,
# variant_probe_coverage.reference_kmer_count,
# self.depth_to_expected_kmer_count(expected_depth + contamination,variant_probe_coverage.alternate_klen),
# self.depth_to_expected_kmer_count((expected_depth + contamination) * self.error_rate / 3,variant_probe_coverage.reference_klen)))
return max(hom_alt_liks)
def hom_alt_lik(self, variant_probe_coverage):
hom_alt_liks = []
# Either alt+cov or alt_covg + contam_covg
for expected_depth in self.expected_depths:
hom_alt_liks.append(
log_lik_R_S_kmer_count(
variant_probe_coverage.alternate_kmer_count,
variant_probe_coverage.reference_kmer_count,
expected_depth, expected_depth * self.error_rate / 3))
for contamination in self.contamination_depths:
hom_alt_liks.append(
log_lik_R_S_kmer_count(
variant_probe_coverage.alternate_kmer_count,
variant_probe_coverage.reference_kmer_count,
expected_depth + contamination,
(expected_depth + contamination) * self.error_rate /
3))
return max(hom_alt_liks)
def het_lik(self, variant_probe_coverage):
# logger.debug("het %s" % variant_probe_coverage.var_name)
if (variant_probe_coverage.alternate_kmer_count +
variant_probe_coverage.reference_kmer_count) == 0:
return MIN_LLK
elif variant_probe_coverage.alternate_percent_coverage < 100 or variant_probe_coverage.reference_percent_coverage < 100:
return MIN_LLK
else:
het_liks = []
for expected_depth in self.expected_depths:
kmer_count_likelihood = log_lik_R_S_kmer_count(
variant_probe_coverage.alternate_kmer_count,
variant_probe_coverage.reference_kmer_count,
self.depth_to_expected_kmer_count(
expected_depth / 2 +
(expected_depth / 2 * self.error_rate / 3),
variant_probe_coverage.alternate_klen),
self.depth_to_expected_kmer_count(
expected_depth / 2 +
(expected_depth / 2 * self.error_rate / 3),
variant_probe_coverage.reference_klen))
ngaps_likelihood_ref = log_lik_probability_of_N_gaps(
expected_depth / 2 +
(expected_depth / 2 * self.error_rate / 3),
variant_probe_coverage.reference_percent_coverage,
self.klen_to_dna_len(
variant_probe_coverage.reference_klen))
ngaps_likelihood_alt = log_lik_probability_of_N_gaps(
expected_depth / 2 +
(expected_depth / 2 * self.error_rate / 3),
variant_probe_coverage.alternate_percent_coverage,
self.klen_to_dna_len(
variant_probe_coverage.alternate_klen))
ngaps_likelihood = ngaps_likelihood_ref + ngaps_likelihood_alt
logger.debug(
"het kmer_count_likelihood: %f %s" %
(kmer_count_likelihood, variant_probe_coverage.var_name))
logger.debug(
"het ngaps_likelihood_ref: %f %s " %
(ngaps_likelihood_ref, variant_probe_coverage.var_name))
logger.debug(
"het ngaps_likelihood_alt: %f %s " %
(ngaps_likelihood_alt, variant_probe_coverage.var_name))
logger.debug(
"het ngaps_likelihood: %f %s " %
(ngaps_likelihood, variant_probe_coverage.var_name))
het_liks.append(kmer_count_likelihood + ngaps_likelihood)
return max(het_liks)
def het_lik(self, variant_probe_coverage):
if (variant_probe_coverage.alternate_kmer_count +
variant_probe_coverage.reference_kmer_count) == 0:
return MIN_LLK
elif variant_probe_coverage.alternate_percent_coverage < 100 or variant_probe_coverage.reference_percent_coverage < 100:
return MIN_LLK
else:
het_liks = []
for expected_depth in self.expected_depths:
het_liks.append(
log_lik_R_S_kmer_count(
variant_probe_coverage.alternate_kmer_count,
variant_probe_coverage.reference_kmer_count,
expected_depth / 2 +
(expected_depth / 2 * self.error_rate / 3),
expected_depth / 2 +
(expected_depth / 2 * self.error_rate / 3)))
return max(het_liks)