gene_copynum_matrix = gene_depth_matrix * 1.0 / (marker_coverages +
(marker_coverages == 0))
#
# convert gene_samples to list:
gene_samples = gene_samples.tolist()
#
# convert gene names to numpy array:
gene_names = numpy.array(gene_names)
#
# indexes for different subject pairs
desired_samples = gene_samples
#
desired_same_sample_idxs, desired_same_subject_idxs, desired_diff_subject_idxs = parse_midas_data.calculate_ordered_subject_pairs(
sample_order_map, desired_samples)
#
snp_sample_idx_map = parse_midas_data.calculate_sample_idx_map(
desired_samples, snp_samples)
gene_sample_idx_map = parse_midas_data.calculate_sample_idx_map(
desired_samples, gene_samples)
#
same_subject_snp_idxs = parse_midas_data.apply_sample_index_map_to_indices(
snp_sample_idx_map, desired_same_subject_idxs)
same_subject_gene_idxs = parse_midas_data.apply_sample_index_map_to_indices(
gene_sample_idx_map, desired_same_subject_idxs)
#
diff_subject_snp_idxs = parse_midas_data.apply_sample_index_map_to_indices(
snp_sample_idx_map, desired_diff_subject_idxs)
diff_subject_gene_idxs = parse_midas_data.apply_sample_index_map_to_indices(
gene_sample_idx_map, desired_diff_subject_idxs)
#
between_host_gene_idxs = [
] # store idxs of genes that change between hosts
sys.stderr.write("Done! (%d genes)\n" % len(reference_genes))
print reference_genes[0:10]
print gene_names[0:10]
# Calculate matrix of number of genes that differ
sys.stderr.write("Calculate gene hamming matrix...\n")
# Either: for all genes in pan-genome
gene_hamming_matrix, num_opportunities = gene_diversity_utils.calculate_coverage_based_gene_hamming_matrix(
gene_depth_matrix, marker_coverages, min_log2_fold_change=4)
#
# Or: just the subset from the MIDAS reference genome
#gene_hamming_matrix = diversity_utils.calculate_coverage_based_gene_hamming_matrix(gene_depth_matrix[reference_gene_idxs,:], marker_coverages, min_log2_fold_change=4)
#
sample_idx_map = parse_midas_data.calculate_sample_idx_map(
high_coverage_samples, samples)
# Calculate which pairs of idxs belong to the same sample, which to the same subject
# and which to different subjects
high_coverage_same_sample_idxs, high_coverage_same_subject_idxs, high_coverage_diff_subject_idxs = parse_midas_data.calculate_subject_pairs(
subject_sample_map, high_coverage_samples)
same_sample_idxs = parse_midas_data.apply_sample_index_map_to_indices(
sample_idx_map, high_coverage_same_sample_idxs)
same_subject_idxs = parse_midas_data.apply_sample_index_map_to_indices(
sample_idx_map, high_coverage_same_subject_idxs)
diff_subject_idxs = parse_midas_data.apply_sample_index_map_to_indices(
sample_idx_map, high_coverage_diff_subject_idxs)
hamming_timepoints = gene_hamming_matrix[same_subject_idxs]
hamming_timepoints.sort()