def _parse_variant_panel(self, row):
allele, median_depth, min_depth, percent_coverage, k_count = self._parse_summary_covgs_row(
row)
params = get_params(allele)
if 'var_name' in params:
var_name = params.get('var_name')
else:
var_name = allele.split('?')[0].split('-')[1]
num_alts = int(params.get("num_alts", 0))
reference_coverages = [
ProbeCoverage(percent_coverage=percent_coverage,
median_depth=median_depth,
min_depth=min_depth,
k_count=k_count)
]
alt_or_ref = 'ref'
alternate_coverages = []
for i in range(num_alts - 1):
row = next(self.reader)
ref_allele, median_depth, min_depth, percent_coverage, k_count = self._parse_summary_covgs_row(
row)
if ref_allele.split('-')[0] != 'ref':
logger.warning("Fewer ref alleles than alt alleles for %s" %
ref_allele)
alternate_coverages.append(
ProbeCoverage(min_depth=min_depth,
k_count=k_count,
percent_coverage=percent_coverage,
median_depth=median_depth))
num_alts -= 1
break
assert ref_allele.split('-')[0] == 'ref'
reference_coverages.append(
ProbeCoverage(percent_coverage=percent_coverage,
median_depth=median_depth,
min_depth=min_depth,
k_count=k_count))
for i in range(num_alts):
row = next(self.reader)
alt_allele, median_depth, min_depth, percent_coverage, k_count = self._parse_summary_covgs_row(
row)
assert alt_allele.split('-')[0] == 'alt'
alternate_coverages.append(
ProbeCoverage(min_depth=min_depth,
k_count=k_count,
percent_coverage=percent_coverage,
median_depth=median_depth))
variant_probe_coverage = VariantProbeCoverage(
reference_coverages=reference_coverages,
alternate_coverages=alternate_coverages,
var_name=var_name,
params=params)
try:
self.variant_covgs[allele].append(variant_probe_coverage)
except KeyError:
self.variant_covgs[allele] = [variant_probe_coverage]
def _name_to_id(self, probe_name):
names = []
params = get_params(probe_name)
if params.get("mut"):
names.append("_".join([params.get("gene"), params.get("mut")]))
var_name = params.get("var_name")
else:
var_name = probe_name.split('?')[0].split('-')[1]
names.append(var_name)
return "-".join(names)
def _parse_seq_panel(self, row):
allele, median_depth, min_depth, percent_coverage, k_count = self._parse_summary_covgs_row(
row)
probe_coverage = ProbeCoverage(percent_coverage=percent_coverage,
median_depth=median_depth,
min_depth=min_depth,
k_count=k_count)
allele_name = allele.split('?')[0]
params = get_params(allele)
panel_type = params.get("panel_type", "presence")
name = params.get('name')
version = params.get('version', '1')
if panel_type in ["variant", "presence"]:
sequence_probe_coverage = SequenceProbeCoverage(
name=name,
probe_coverage=probe_coverage,
version=version,
length=params.get("length"))
try:
self.covgs[panel_type][name][version] = sequence_probe_coverage
except KeyError:
self.covgs[panel_type][name] = {}
self.covgs[panel_type][name][version] = sequence_probe_coverage
else:
# Species panels are treated differently
l = int(params.get("length", -1))
try:
self.covgs[panel_type][name]["total_bases"] += l
if percent_coverage > 75 and median_depth > 0:
self.covgs[panel_type][name]["percent_coverage"].append(
percent_coverage)
self.covgs[panel_type][name]["length"].append(l)
self.covgs[panel_type][name]["median"].append(median_depth)
except KeyError:
if panel_type not in self.covgs:
self.covgs[panel_type] = {}
self.covgs[panel_type][name] = {}
self.covgs[panel_type][name]["total_bases"] = l
if percent_coverage > 75 and median_depth > 0:
self.covgs[panel_type][name]["percent_coverage"] = [
percent_coverage
]
self.covgs[panel_type][name]["length"] = [l]
self.covgs[panel_type][name]["median"] = [median_depth]
else:
self.covgs[panel_type][name]["percent_coverage"] = []
self.covgs[panel_type][name]["length"] = []
self.covgs[panel_type][name]["median"] = []
def _create_variant(self, probe_name):
names = []
params = get_params(probe_name)
if params.get("mut"):
names.append("_".join([params.get("gene"), params.get("mut")]))
var_name = probe_name.split('?')[0].split('-')[1]
names.append(var_name)
try:
# If it's a variant panel we can create a variant
ref, start, alt = split_var_name(var_name)
return Variant.create(start=start,
reference_bases=ref,
alternate_bases=[alt],
names=names,
info=params)
except AttributeError:
return None
def run(parser, args):
genes = {}
skip_list = {
"tem": ["191", "192"],
"oxa": ["12", "14", "33"],
"shv": ["12", "6"]
}
check_args(args)
if args.seq:
build_binary()
if args.also_genotype:
_out_dict = run_genotype(parser, args)
else:
_out_dict = {}
_out_dict[args.sample] = {}
_out_dict[args.sample]["paths"] = {}
out_dict = _out_dict[args.sample]["paths"]
wb = WebServer(port=0,
args=[args.ctx],
memory=args.memory,
mccortex_path=args.mccortex31_path)
logger.debug("Loading binary")
wb.start()
logger.debug("Walking the graph")
gw = GraphWalker(proc=wb.mccortex, kmer_size=args.kmer, print_depths=True)
with open(args.probe_set, 'r') as infile:
for i, record in enumerate(SeqIO.parse(infile, "fasta")):
repeat_kmers = get_repeat_kmers(record, args.kmer)
params = get_params(record.id)
gene_name = params.get("name", i)
version = params.get("version", i)
if gene_name not in genes:
logger.debug("Loading kmer data for %s" % (gene_name))
last_kmer = str(record.seq)[-args.kmer:]
start_kmer, skipped = find_start_kmer(str(record.seq), gw.mcq,
args.kmer)
if gene_name not in genes:
genes[gene_name] = {}
genes[gene_name]["pathdetails"] = []
genes[gene_name]["known_kmers"] = ""
if version not in skip_list.get(gene_name, []) and start_kmer:
pd = PathDetails(start_kmer,
last_kmer,
len(record.seq),
skipped=skipped,
v=version)
pd.set_repeat_kmers(repeat_kmers)
genes[gene_name]["pathdetails"].append(pd)
if gene_name in genes:
genes[gene_name]["known_kmers"] += "%sN" % str(record.seq)
for gene_name, gene_dict in genes.items():
logger.debug("Walking graph with seeds defined by %s" % gene_name)
paths = get_paths_for_gene(gene_name, gene_dict, gw)
if args.show_all_paths:
out_dict[gene_name] = paths.values()
else:
if len(paths.keys()) > 1:
# choose best version
best_path = choose_best_assembly(paths.values())
elif len(paths.keys()) == 1:
best_path = paths.values()[0]
else:
best_path = {"found": False}
out_dict[gene_name] = [best_path]
print(json.dumps(_out_dict, sort_keys=False, indent=4))
logger.info("Cleaning up")
if wb is not None:
wb.stop()