def run(params):
with open(params.multisample_profile_path, "r") as profile_file:
multisample_profile = json.load(profile_file)
count_table = generate_table_with_anchor_counts(multisample_profile["Counts"])
logging.info("Loaded %i regions", len(count_table))
logging.info("Normalizing counts")
sample_stats = multisample_profile["Parameters"]
common.depth_normalize_counts(sample_stats, count_table)
logging.info("Filtering counts")
count_table = common.filter_counts_by_magnitude(
count_table, params.min_inrepeat_reads
)
if params.target_region_path:
target_regions = load_target_regions(params.target_region_path)
logging.info("Restricting analysis to %i regions", len(target_regions))
count_table = common.filter_counts_by_region(count_table, target_regions)
logging.info("%i regions left after filtering", len(count_table))
manifest = common.load_manifest(params.manifest_path)
sample_status = common.extract_case_control_assignments(manifest)
logging.info("Comparing counts")
common.compare_counts(params.test_params, sample_status, count_table)
logging.info("Correcting p-values")
common.correct_pvalues(count_table)
output_results(count_table, params.output_path)
logging.info("Done")
def run(params):
with open(params.multisample_profile_path, "r") as profile_file:
multisample_profile = json.load(profile_file)
count_table = generate_table_with_anchor_counts(
multisample_profile["Counts"])
logging.info("Loaded %i regions", len(count_table))
logging.info("Normalizing counts")
sample_stats = multisample_profile["Parameters"]
common.depth_normalize_counts(sample_stats, count_table)
if params.target_region_path:
target_regions = load_target_regions(params.target_region_path)
logging.info("Restricting analysis to %i regions", len(target_regions))
count_table = common.filter_counts_by_region(count_table,
target_regions)
manifest = common.load_manifest(params.manifest_path)
sample_status = common.extract_case_control_assignments(manifest)
header = "contig\tstart\tend\tmotif\ttop_case_zscore\thigh_case_counts\tcounts"
with open(params.output_path, "w") as results_file:
print(header, file=results_file)
for row in count_table:
region_encoding = row["region"]
if region_encoding == "unaligned":
continue
contig, coords = region_encoding.rsplit(":", 1)
start, end = coords.split("-")
start, end = int(start), int(end)
top_case_zscore, cases_with_high_counts = common.run_zscore_analysis(
sample_status, row["sample_counts"])
if len(cases_with_high_counts) == 0:
continue
encoded_case_info = ",".join(
"{}:{:.2f}".format(s, c)
for s, c in cases_with_high_counts.items())
count_encoding = ",".join(
["{:.2f}".format(c) for _, c in row["sample_counts"].items()])
print(
contig,
start,
end,
row["unit"],
"{:.2f}".format(top_case_zscore),
encoded_case_info,
count_encoding,
sep="\t",
file=results_file,
)
logging.info("Done")
def main():
common.init_logger()
parameters = load_parameters()
samples = common.load_manifest(parameters['manifest_path'])
sample_depths = {}
combined_regions = process_samples(parameters, samples, sample_depths)
normalize_counts(sample_depths, combined_regions)
output_json = create_json(combined_regions)
write_json(parameters, output_json)
def main():
common.init_logger()
parameters = load_parameters()
combined_counts = common.load_combined_json(parameters['counts_path'])
samples = common.load_manifest(parameters['manifest_path'])
count_table = generate_count_table(combined_counts)
count_table = common.filter_counts(parameters['min_count'], count_table)
sample_status = common.extract_case_control_assignments(samples)
common.compare_counts(sample_status, count_table)
common.correct_pvalues(count_table)
output_results(count_table, parameters['output_path'])
logging.info('Done')
def main():
common.init_logger()
parameters = load_parameters()
combined_counts = common.load_combined_json(parameters['counts_path'])
samples = common.load_manifest(parameters['manifest_path'])
target_regions = None
if parameters['target_regions']:
target_regions = load_target_regions(parameters['target_regions'])
count_table = generate_count_table(combined_counts)
count_table = common.filter_counts(count_table, parameters['min_count'],
target_regions)
logging.info('%i regions left after initial filtering', len(count_table))
sample_status = common.extract_case_control_assignments(samples)
common.compare_counts(parameters['test_params'], sample_status,
count_table)
common.correct_pvalues(count_table)
output_results(count_table, parameters['output_path'])
logging.info('Done')
def run(params):
with open(params.multisample_profile_path, "r") as profile_file:
multisample_profile = json.load(profile_file)
count_table = common.generate_table_with_irr_pair_counts(
multisample_profile["Counts"])
logging.info("Loaded %i regions", len(count_table))
logging.info("Normalizing counts")
sample_stats = multisample_profile["Parameters"]
common.depth_normalize_counts(sample_stats, count_table)
manifest = common.load_manifest(params.manifest_path)
sample_status = common.extract_case_control_assignments(manifest)
header = "motif\ttop_case_zscore\thigh_case_counts\tcounts"
with open(params.output_path, "w") as results_file:
print(header, file=results_file)
for row in count_table:
top_case_zscore, cases_with_high_counts = common.run_zscore_analysis(
sample_status, row["sample_counts"])
if len(cases_with_high_counts) == 0:
continue
encoded_case_info = ",".join(
"{}:{:.2f}".format(s, c)
for s, c in cases_with_high_counts.items())
count_encoding = ",".join(
["{:.2f}".format(c) for _, c in row["sample_counts"].items()])
motif = row["unit"]
print(
motif,
"{:.2f}".format(top_case_zscore),
encoded_case_info,
count_encoding,
sep="\t",
file=results_file,
)