same_sample_idxs, same_subject_idxs, diff_subject_idxs = parse_midas_data.calculate_subject_pairs(
subject_sample_map, snp_samples)
#
diff_subject_snp_plowers = []
diff_subject_snp_puppers = []
diff_subject_gene_plowers = []
diff_subject_gene_puppers = []
between_host_gene_idx_map = {}
low_divergence_between_host_gene_idx_map = {}
for sample_pair_idx in xrange(0, len(diff_subject_idxs[0])):
#
snp_i = diff_subject_idxs[0][sample_pair_idx]
snp_j = diff_subject_idxs[1][sample_pair_idx]
#
plower, pupper = stats_utils.calculate_poisson_rate_interval(
snp_difference_matrix[snp_i, snp_j],
snp_opportunity_matrix[snp_i, snp_j],
alpha=0.05)
#
diff_subject_snp_plowers.append(plower)
diff_subject_snp_puppers.append(pupper)
#
# clip lower bounds
diff_subject_snp_plowers = numpy.clip(diff_subject_snp_plowers, 1e-07,
1e09)
# Sort all four lists by ascending lower bound on SNP changes, then gene changes
diff_subject_snp_plowers, diff_subject_snp_puppers = (
numpy.array(x) for x in zip(
*sorted(zip(diff_subject_snp_plowers, diff_subject_snp_puppers))))
##############################################################################
#
#print typical_same_subject_snp_opportunities, typical_diff_subject_snp_opportunities
#print typical_same_subject_gene_opportunities, typical_diff_subject_gene_opportunities
Lsnps = typical_diff_subject_snp_opportunities
Lgenes = typical_diff_subject_gene_opportunities
same_subject_snp_plowers = []
same_subject_snp_puppers = []
same_subject_gene_plowers = []
same_subject_gene_puppers = []
for sample_pair_idx in xrange(0, len(same_subject_snp_idxs[0])):
i = same_subject_snp_idxs[0][sample_pair_idx]
j = same_subject_snp_idxs[1][sample_pair_idx]
plower, pupper = stats_utils.calculate_poisson_rate_interval(
snp_difference_matrix[i, j], snp_opportunity_matrix[i, j], alpha)
same_subject_snp_plowers.append(plower)
same_subject_snp_puppers.append(pupper)
snp_differences = diversity_utils.calculate_snp_differences_between(
i, j, allele_counts_map, passed_sites_map, min_change=min_change)
i = same_subject_gene_idxs[0][sample_pair_idx]
j = same_subject_gene_idxs[1][sample_pair_idx]
gene_differences = gene_diversity_utils.calculate_gene_differences_between(
i, j, gene_depth_matrix, marker_coverages, min_log2_fold_change=4)
plower, pupper = stats_utils.calculate_poisson_rate_interval(
gene_difference_matrix[i, j], gene_opportunity_matrix[i, j])
median_depths = []
depth_lowers = []
depth_uppers = []
# Now do polymorphism axis
for sample in highcoverage_samples:
within_sites, between_sites, total_sites = sfs_utils.calculate_polymorphism_rates_from_sfs_map(
sfs_map[sample])
within_rate = within_sites * 1.0 / total_sites
between_rate = between_sites * 1.0 / total_sites
between_rates.append(between_rate)
within_rates.append(within_rate)
# Calculate 95% confidence intervals
within_rate_lower, within_rate_upper = stats_utils.calculate_poisson_rate_interval(
within_sites, total_sites, alpha=0.05)
within_rate_lowers.append(within_rate_lower)
within_rate_uppers.append(within_rate_upper)
depths, counts = sfs_utils.calculate_depth_distribution_from_sfs_map(
sfs_map[sample])
dlower, dupper = stats_utils.calculate_IQR_from_distribution(
depths, counts)
dmedian = stats_utils.calculate_median_from_distribution(
depths, counts)
depth_lowers.append(dlower)
depth_uppers.append(dupper)
median_depths.append(dmedian)
sample_names.append(sample)
for sample_pair_idx in xrange(0, len(same_subject_snp_idxs[0])):
i = same_subject_snp_idxs[0][sample_pair_idx]
j = same_subject_snp_idxs[1][sample_pair_idx]
i2 = desired_same_subject_idxs[0][sample_pair_idx]
j2 = desired_same_subject_idxs[1][sample_pair_idx]
if pis[i2] > pis[j2]:
pi_idx = i2
else:
pi_idx = j2
# Calculate max pi between two samples
plower, pupper = stats_utils.calculate_poisson_rate_interval(
total_pis[pi_idx], total_pi_opportunities[pi_idx], alpha)
pmid = total_pis[pi_idx] * 1.0 / total_pi_opportunities[pi_idx]
same_subject_pi_plowers.append(plower)
same_subject_pi_pmids.append(pmid)
same_subject_pi_puppers.append(pupper)
# Calculate SNP changes
plower, pupper = stats_utils.calculate_poisson_rate_interval(
snp_difference_matrix[i, j], snp_opportunity_matrix[i, j], alpha)
pmid = snp_difference_matrix[i, j] * 1.0 / snp_opportunity_matrix[i, j]
same_subject_snp_plowers.append(plower)
same_subject_snp_pmids.append(pmid)
same_subject_snp_puppers.append(pupper)