def main():
args = parse_args()
bam_map = parse_bam_map_file(args.bam_map_file)
vcf_reader = vcf.Reader(args.vcf_file)
vep_cols = parse_vep_cols(vcf_reader)
gene_list = parse_genes(args.genes)
records_per_gene = parse_vcf_file(vcf_reader, vep_cols, gene_list)
for gene, records in records_per_gene.items():
gene_dir = os.path.join(args.output_dir, gene)
if not os.path.exists(gene_dir):
os.mkdir(gene_dir)
# Create script file
with open(os.path.join(gene_dir, SCRIPT_NAME), "w") as sf:
sf.write(generate_autoigv_cmd(args.python, args.autoigv, args.genome))
# Create positions file
with open(os.path.join(gene_dir, POSITIONS_NAME), "w") as pf:
pf.write(generate_autoigv_positions(gene, records, bam_map))
# Create master script file
with open(os.path.join(args.output_dir, "run_all.sh"), "w") as mf:
mf.write(generate_master_script())
# Create prefs file
with open(os.path.join(args.output_dir, PREFS_NAME), "w") as pf:
pf.write(generate_prefs())
def main():
args = parse_args()
bam_map = parse_bam_map_file(args.bam_map_file)
vcf_reader = vcf.Reader(args.vcf_file)
vep_cols = parse_vep_cols(vcf_reader)
gene_list = parse_genes(args.genes)
records_per_gene = parse_vcf_file(vcf_reader, vep_cols, gene_list)
for gene, records in records_per_gene.items():
gene_dir = os.path.join(args.output_dir, gene)
if not os.path.exists(gene_dir):
os.mkdir(gene_dir)
# Create script file
with open(os.path.join(gene_dir, SCRIPT_NAME), "w") as sf:
sf.write(
generate_autoigv_cmd(args.python, args.autoigv, args.genome))
# Create positions file
with open(os.path.join(gene_dir, POSITIONS_NAME), "w") as pf:
pf.write(generate_autoigv_positions(gene, records, bam_map))
# Create master script file
with open(os.path.join(args.output_dir, "run_all.sh"), "w") as mf:
mf.write(generate_master_script())
# Create prefs file
with open(os.path.join(args.output_dir, PREFS_NAME), "w") as pf:
pf.write(generate_prefs())
def main():
"""Main program"""
# Argument parsing
args = parse_args()
# Setup
vcf_reader = vcf.Reader(args.input_vcf)
vep_cols = parse_vep_cols(vcf_reader)
# Create set of genes to be excluded
excl_genes_set = build_exclude_genes(args.exclude_genes)
# Create set of positions to be excluded
excl_pos_set = build_exclude_positions(args.exclude_positions)
# Build dict of genes with affected samples
# Sets: num_samples, num_samples_mod_impact, num_samples_high_impact
SampleSets = namedtuple("SampleSets", ["all", "moderate", "high"])
genes = defaultdict(lambda: SampleSets(set(), set(), set()))
# Iterate over VCF file
for record in vcf_reader:
# Filter on position, if applicable
pos_id = create_pos_id(record.CHROM, record.POS)
if pos_id in excl_pos_set:
continue
# Filter on NUM_SAMPLES
if args.max_samples and record.INFO["NUM_SAMPLES"] > args.max_samples:
continue
# Parse VEP output and select the first and only one
vep_effect = parse_vep(vep_cols, record, tag="TOP_CSQ")[0]
# Skip if symbol is absent
if args.symbol and vep_effect["SYMBOL"] == "":
continue
# Exclude on gene ID or symbol
if vep_effect["Gene"] in excl_genes_set or vep_effect[
"SYMBOL"] in excl_genes_set:
continue
# Extract gene ID and symbol
gid, gsymbol = vep_effect["Gene"], vep_effect["SYMBOL"]
# Extract calls with minimum depth
calls = []
for call in record.samples:
depth = getattr(call.data, "DP", 0)
allele_depths = getattr(call.data, "AD", (0, 0))
if (call.gt_type != 0 and depth >= args.min_depth
and allele_depths[1] /
(allele_depths[0] + allele_depths[1]) <
args.homo_vaf_threshold):
calls.append(call)
# Extract samples
samples = set(c.sample for c in calls)
# Add samples to genes dict; using gid and gsymbol for readability
genes[(gid, gsymbol)].all.update(samples)
# Update sample lists based on variant type
if any([eff in vep_effect["Consequence"] for eff in HIGH_IMPACT]):
genes[(gid, gsymbol)].high.update(samples)
elif any([eff in vep_effect["Consequence"]
for eff in MODERATE_IMPACT]):
genes[(gid, gsymbol)].moderate.update(samples)
# Order genes by number of affected samples
genes_list = [(gene[0], gene[1], len(sets[0]), len(sets[1]), len(sets[2]))
for gene, sets in genes.items()]
genes_list.sort(key=lambda x: x[2], reverse=True)
# Output sorted gene list
header = "\t".join([
"gene_id", "gene_symbol", "num_samples",
"num_samples_with_moderate_effect", "num_samples_with_high_effect"
]) + "\n"
args.output.write(header)
for gene in genes_list:
line = "\t".join(map(str, gene)) + "\n"
args.output.write(line)
# Cleanup
args.output.close()
def main():
"""Main program"""
# Argument parsing
args = parse_args()
# Setup
vcf_reader = vcf.Reader(args.input_vcf)
vep_cols = parse_vep_cols(vcf_reader)
# Create set of genes to be excluded
excl_genes_set = build_exclude_genes(args.exclude_genes)
# Create set of positions to be excluded
excl_pos_set = build_exclude_positions(args.exclude_positions)
# Build dict of genes with affected samples
# Sets: num_samples, num_samples_mod_impact, num_samples_high_impact
SampleSets = namedtuple("SampleSets", ["all", "moderate", "high"])
genes = defaultdict(lambda: SampleSets(set(), set(), set()))
# Iterate over VCF file
for record in vcf_reader:
# Filter on position, if applicable
pos_id = create_pos_id(record.CHROM, record.POS)
if pos_id in excl_pos_set:
continue
# Filter on NUM_SAMPLES
if args.max_samples and record.INFO["NUM_SAMPLES"] > args.max_samples:
continue
# Parse VEP output and select the first and only one
vep_effect = parse_vep(vep_cols, record, tag="TOP_CSQ")[0]
# Skip if symbol is absent
if args.symbol and vep_effect["SYMBOL"] == "":
continue
# Exclude on gene ID or symbol
if vep_effect["Gene"] in excl_genes_set or vep_effect["SYMBOL"] in excl_genes_set:
continue
# Extract gene ID and symbol
gid, gsymbol = vep_effect["Gene"], vep_effect["SYMBOL"]
# Extract calls with minimum depth
calls = []
for call in record.samples:
depth = getattr(call.data, "DP", 0)
allele_depths = getattr(call.data, "AD", (0, 0))
if (call.gt_type != 0 and depth >= args.min_depth and allele_depths[1] / (allele_depths[0] + allele_depths[1]) < args.homo_vaf_threshold):
calls.append(call)
# Extract samples
samples = set(c.sample for c in calls)
# Add samples to genes dict; using gid and gsymbol for readability
genes[(gid, gsymbol)].all.update(samples)
# Update sample lists based on variant type
if any([eff in vep_effect["Consequence"] for eff in HIGH_IMPACT]):
genes[(gid, gsymbol)].high.update(samples)
elif any([eff in vep_effect["Consequence"] for eff in MODERATE_IMPACT]):
genes[(gid, gsymbol)].moderate.update(samples)
# Order genes by number of affected samples
genes_list = [(gene[0], gene[1], len(sets[0]), len(sets[1]), len(sets[2])) for gene, sets in genes.items()]
genes_list.sort(key=lambda x: x[2], reverse=True)
# Output sorted gene list
header = "\t".join(["gene_id", "gene_symbol", "num_samples", "num_samples_with_moderate_effect", "num_samples_with_high_effect"]) + "\n"
args.output.write(header)
for gene in genes_list:
line = "\t".join(map(str, gene)) + "\n"
args.output.write(line)
# Cleanup
args.output.close()