def prepare_resource(datadir, parameters):
genome = parameters.genome
region_file = os.path.join(datadir, "CYP2D6_region_%s.bed" % genome)
snp_file = os.path.join(datadir, "CYP2D6_SNP_%s.txt" % genome)
gmm_file = os.path.join(datadir, "CYP2D6_gmm.txt")
star_table = os.path.join(datadir, "star_table.txt")
variant_file = os.path.join(datadir,
"CYP2D6_target_variant_%s.txt" % genome)
variant_homology_file = os.path.join(
datadir, "CYP2D6_target_variant_homology_region_%s.txt" % genome)
haplotype_file = os.path.join(datadir, "CYP2D6_haplotype_%s.txt" % genome)
star_combinations = get_hap_table(star_table)
for required_file in [
region_file,
snp_file,
variant_file,
variant_homology_file,
haplotype_file,
gmm_file,
]:
if os.path.exists(required_file) == 0:
raise Exception("File %s not found." % required_file)
snp_db = get_snp_position(snp_file)
var_db = get_snp_position(variant_file)
var_homo_db = get_snp_position(variant_homology_file)
haplotype_db = {}
for variant in HAPLOTYPE_VAR:
haplotype_db.setdefault(variant,
get_snp_position(haplotype_file, variant))
var_list = []
with open(variant_file) as f:
for line in f:
if line[0] != "#":
var_name = line.split()[-1]
var_list.append(var_name)
with open(variant_homology_file) as f:
for line in f:
if line[0] != "#":
var_name = line.split()[-1]
var_list.append(var_name)
gmm_parameter = parse_gmm_file(gmm_file)
region_dic = parse_region_file(region_file)
call_parameters = resource_info(
genome,
gmm_parameter,
region_dic,
snp_db,
var_db,
var_homo_db,
haplotype_db,
var_list,
star_combinations,
)
return call_parameters
def main():
parameters = load_parameters()
manifest = parameters.manifest
outdir = parameters.outDir
genome = parameters.genome
prefix = parameters.prefix
reference_fasta = parameters.reference
threads = parameters.threads
path_count_file = parameters.countFilePath
logging.basicConfig(level=logging.DEBUG)
# Prepare data files
datadir = os.path.join(os.path.dirname(__file__), "data")
region_file = os.path.join(datadir, "CYP2D6_region_%s.bed" % genome)
snp_file = os.path.join(datadir, "CYP2D6_SNP_%s.txt" % genome)
gmm_file = os.path.join(datadir, "CYP2D6_gmm.txt")
table_path = "full_star_table"
if parameters.knownFunction:
table_path = "known_function_star_table"
if parameters.includeNewStar:
table_path = "include_new_star_table"
star_table = os.path.join(datadir, table_path, "star_table.txt")
variant_file = os.path.join(datadir, table_path,
"CYP2D6_target_variant_%s.txt" % genome)
variant_homology_file = os.path.join(
datadir, table_path,
"CYP2D6_target_variant_homology_region_%s.txt" % genome)
star_combinations = get_hap_table(star_table)
for required_file in [
region_file,
snp_file,
variant_file,
variant_homology_file,
gmm_file,
]:
if os.path.exists(required_file) == 0:
raise Exception("File %s not found." % required_file)
if os.path.exists(outdir) == 0:
os.makedirs(outdir)
snp_db = get_snp_position(snp_file)
var_db = get_snp_position(variant_file)
var_homo_db = get_snp_position(variant_homology_file)
var_list = []
with open(variant_file) as f:
for line in f:
if line[0] != "#":
var_name = line.split()[-1]
var_list.append(var_name)
with open(variant_homology_file) as f:
for line in f:
if line[0] != "#":
var_name = line.split()[-1]
var_list.append(var_name)
gmm_parameter = parse_gmm_file(gmm_file)
region_dic = parse_region_file(region_file)
resource_info = namedtuple(
"resource_info",
"genome gmm_parameter region_dic snp_db var_db var_homo_db var_list star_combinations",
)
call_parameters = resource_info(
genome,
gmm_parameter,
region_dic,
snp_db,
var_db,
var_homo_db,
var_list,
star_combinations,
)
out_json = os.path.join(outdir, prefix + ".json")
out_tsv = os.path.join(outdir, prefix + ".tsv")
final_output = {}
with open(manifest) as read_manifest:
for line in read_manifest:
bam_name = line.strip()
sample_id = os.path.splitext(os.path.basename(bam_name))[0]
count_file = None
if path_count_file is not None:
count_file = os.path.join(path_count_file,
sample_id + "_count.txt")
if os.path.exists(bam_name) == 0 or (count_file is not None
and os.path.exists(count_file)
== 0):
logging.warning("Input file for sample %s does not exist.",
sample_id)
else:
logging.info("Processing sample %s at %s", sample_id,
datetime.datetime.now())
cyp2d6_call = d6_star_caller(bam_name, call_parameters,
threads, count_file,
reference_fasta)._asdict()
# Use normalized coverage MAD across stable regions
# as a sample QC measure.
if cyp2d6_call["Coverage_MAD"] > MAD_THRESHOLD:
logging.warning(
"Sample %s has uneven coverage. CN calls may be unreliable.",
sample_id,
)
final_output.setdefault(sample_id, cyp2d6_call)
# Write to json
logging.info("Writing to json at %s", datetime.datetime.now())
with open(out_json, "w") as json_output:
json.dump(final_output, json_output)
# Write to tsv
logging.info("Writing to tsv at %s", datetime.datetime.now())
header = ["Sample", "Genotype", "Filter"]
with open(out_tsv, "w") as tsv_output:
tsv_output.write("\t".join(header) + "\n")
for sample_id in final_output:
final_call = final_output[sample_id]
output_per_sample = [
sample_id,
final_call["Genotype"],
final_call["Filter"],
]
tsv_output.write("\t".join(str(a)
for a in output_per_sample) + "\n")
def main():
parameters = load_parameters()
manifest = parameters.manifest
outdir = parameters.outDir
genome = parameters.genome
prefix = parameters.prefix
threads = parameters.threads
reference_fasta = parameters.reference
path_count_file = parameters.countFilePath
logging.basicConfig(level=logging.DEBUG)
datadir = os.path.join(os.path.dirname(__file__), "data")
# Region file to use
region_file = os.path.join(datadir, "SMN_region_%s.bed" % genome)
snp_file = os.path.join(datadir, "SMN_SNP_%s.txt" % genome)
variant_file = os.path.join(datadir, "SMN_target_variant_%s.txt" % genome)
gmm_file = os.path.join(datadir, "SMN_gmm.txt")
for required_file in [region_file, snp_file, variant_file, gmm_file]:
if os.path.exists(required_file) == 0:
raise Exception('File %s not found.' % required_file)
if os.path.exists(outdir) == 0:
os.makedirs(outdir)
snp_db = get_snp_position(snp_file)
variant_db = get_snp_position(variant_file)
gmm_parameter = parse_gmm_file(gmm_file)
region_dic = parse_region_file(region_file)
out_json = os.path.join(outdir, prefix + '.json')
out_tsv = os.path.join(outdir, prefix + '.tsv')
final_output = {}
with open(manifest) as read_manifest:
for line in read_manifest:
bam_name = line.strip()
sample_id = os.path.splitext(os.path.basename(bam_name))[0]
count_file = None
if path_count_file is not None:
count_file = os.path.join(
path_count_file, sample_id + '_count.txt')
if count_file is None and os.path.exists(bam_name) == 0:
logging.warning(
'Input alignmet file for sample %s does not exist.', sample_id)
elif count_file is not None and os.path.exists(count_file) == 0:
logging.warning(
'Input count file for sample %s does not exist', sample_id)
else:
logging.info(
'Processing sample %s at %s', sample_id,
datetime.datetime.now()
)
smn_call = smn_cn_caller(
bam_name, region_dic, gmm_parameter,
snp_db, variant_db, threads, count_file, reference_fasta
)
# Use normalized coverage MAD across stable regions
# as a sample QC measure.
if smn_call['Coverage_MAD'] > MAD_THRESHOLD:
logging.warning(
"Sample %s has uneven coverage. CN calls may be \
unreliable.", sample_id)
final_output.setdefault(sample_id, smn_call)
# Write to json
logging.info('Writing to json at %s', datetime.datetime.now())
with open(out_json, 'w') as json_output:
json.dump(final_output, json_output)
# Write to tsv
logging.info('Writing to tsv at %s', datetime.datetime.now())
write_to_tsv(final_output, out_tsv)