def run(args):
if os.path.exists(args.output_file):
logging.info("Output %s exists. Nope", args.output_file)
return
results_order = []
results = {}
logging.info("Streaming file for groups")
for i,line in Utilities.iterate_file(args.input_file):
if i==0: continue
comps = line.strip().split()
key = comps[0]
if not key in results:
results_order.append(key)
results[key] = 0
logging.log(9, "Key: %s", str(key))
results[key] += 1
r = []
logging.info("Producing output")
for key in results_order:
r.append((key, results[key]))
r = pandas.DataFrame(r, columns=["key","count"])
logging.info("Saving")
Utilities.ensure_requisite_folders(args.output_file)
Utilities.save_dataframe(r, args.output_file)
logging.info("Finished.")
def _get_metadata(path, index):
m = []
for i, line in Utilities.iterate_file(path):
if i == 0: continue
comps = line.strip().split()
chr = "chr" + comps[0]
pos = int(comps[1])
variant = comps[2]
non_effect = comps[3]
effect = comps[4]
frequency = float(comps[5])
if chr in index and pos in index[chr]:
m.append((variant, chr, pos, non_effect, effect, frequency))
return m
def _get_metadata(path, index, extra_cols=None):
m = []
for i, line in Utilities.iterate_file(path):
if i == 0: continue
comps = line.strip().split()
chr = "chr" + comps[0]
pos = int(comps[1])
variant = comps[2]
non_effect = comps[3]
effect = comps[4]
frequency = float(comps[5]) if comps[5] != "NA" else numpy.nan
extra_entries = []
for ecol in extra_cols:
extra_entries.append(comps[ecol])
if chr in index and pos in index[chr]:
m.append([variant, chr, pos, non_effect, effect, frequency] +
extra_entries)
return m
def input_generator(input_file, samples):
introns = set()
for i,line in Utilities.iterate_file(input_file):
comps = line.strip().split()
if i==0:
_index = [i for i,x in enumerate(comps) if x in samples]
yield _to_line(comps, _index, "NAME")
continue
name = comps[3]
name_ = name.split(":")
name = "_".join(["intron", name_[0].split("chr")[1], name_[1], name_[2]])
gene = name_[4]
if name in introns:
continue
introns.add(name)
yield _to_line(comps, _index, name)
def run(args):
logging.info("Starting process")
vf = pq.ParquetFile(args.parquet_genotype_metadata)
m = None
last_chromosome = None
r = []
for i, line in Utilities.iterate_file(args.regions):
if i == 0: continue
comps = line.strip().split()
count, m, last_chromosome = count_variants(comps[0], comps[1],
comps[2], vf, m,
last_chromosome, args)
r.append((comps[0], comps[1], comps[2], count))
r = Utilities.to_dataframe(r, ["chromosome", "start", "end", "count"])
Utilities.save_dataframe(r, args.output)
logging.info("Finished process")
def run(args):
r = re.compile(args.pattern)
files = [x for x in os.listdir(args.folder) if r.search(x)]
if args.sort_groups:
files = sorted(files, key=lambda x: _key(x, r, args.sort_groups))
output_firstline = True
Utilities.ensure_requisite_folders(args.output)
logging.info("Starting concatenation")
with gzip.open(args.output, "w") as o:
for file in files:
path = os.path.join(args.folder, file)
logging.log(9, "Opening %s", path)
for i, line in Utilities.iterate_file(path):
if i==0:
if output_firstline:
o.write(line.encode())
if not args.headerless:
output_firstline = False
continue
o.write(line.encode())
logging.info("Finished")
