def load_samples(args):
s = None
if args.text_sample_ids:
if len(args.text_sample_ids) == 1:
s = pandas.read_table(args.text_sample_ids[0],
header=None,
names=["FID", "IID"])
elif args.text_sample_ids[1] == "UKB":
k = pandas.read_table(args.text_sample_ids[0], sep=" ")
k = k[k.sex != "D"].reset_index(drop=True)
s = k[["ID_1", "ID_2"]].rename(columns={
"ID_1": "FID",
"ID_2": "IID"
})
elif args.vcf_genotypes:
from metax.genotype import CYVCF2Genotype
s = CYVCF2Genotype.get_samples(args.vcf_genotypes[0])
elif args.bgen_genotypes:
from metax.genotype import BGENGenotype
s = BGENGenotype.get_samples(args.bgen_genotypes[0])
elif args.generate_sample_ids:
s = ["ID_{}".format(x) for x in range(0, args.generate_sample_ids)]
s = [(x, x) for x in s]
s = pandas.DataFrame(data=s, columns=["FID", "IID"])
if s is None:
raise Exceptions.InvalidArguments("Unsupported samples argument")
return s
def dosage_generator(args, variant_mapping=None, weights=None):
if args.liftover:
logging.info("Acquiring liftover conversion")
liftover_chain = pyliftover.LiftOver(args.liftover)
liftover_conversion = lambda chr, pos: Genomics.lift(
liftover_chain, chr, pos, args.zero_based_positions)
else:
liftover_chain = None
liftover_conversion = None
whitelist = None
if variant_mapping and type(variant_mapping) == dict:
logging.info("Setting whitelist from mapping keys")
whitelist = set(variant_mapping.keys())
else:
logging.info("Setting whitelist from available models")
whitelist = set(weights.rsid)
d = None
if args.text_genotypes:
from metax.genotype import DosageGenotype
d = DosageGenotype.dosage_files_geno_lines(
args.text_genotypes,
variant_mapping=variant_mapping,
whitelist=whitelist,
skip_palindromic=args.skip_palindromic,
liftover_conversion=liftover_conversion)
elif args.bgen_genotypes:
from metax.genotype import BGENGenotype
d = BGENGenotype.bgen_files_geno_lines(
args.bgen_genotypes,
variant_mapping=variant_mapping,
force_colon=args.force_colon,
use_rsid=args.bgen_use_rsid,
whitelist=whitelist,
skip_palindromic=args.skip_palindromic)
elif args.vcf_genotypes:
from metax.genotype import CYVCF2Genotype
d = CYVCF2Genotype.vcf_files_geno_lines(
args.vcf_genotypes,
mode=args.vcf_mode,
variant_mapping=variant_mapping,
whitelist=whitelist,
skip_palindromic=args.skip_palindromic,
liftover_conversion=liftover_conversion)
if d is None:
raise Exceptions.InvalidArguments("unsupported genotype input")
if args.force_mapped_metadata:
d = Genotype.force_mapped_metadata(d, args.force_mapped_metadata)
return d