def interval2genes(interval, skip=('CGH', '-')):
"""Squash intervals into named genes."""
curr_name = None
curr_chrom = None
curr_start = None
curr_end = None
curr_len = 0
for chrom, start, end, name in ngfrills.parse_regions(interval):
if name in skip:
continue
if chrom != curr_chrom or name != curr_name:
if curr_name is not None:
# yield (curr_chrom, curr_start, curr_end, curr_name)
if curr_len > 1:
# Emit
yield (curr_chrom, curr_start, curr_end, curr_name)
else:
print("Single-probe gene is probably CGH:", curr_name,
file=sys.stderr)
# Reset
curr_name = name
curr_chrom = chrom
curr_start = start
curr_len = 0
# Extend
curr_end = end
curr_len += 1
if curr_name is not None:
if curr_len > 1:
# Emit
yield (curr_chrom, curr_start, curr_end, curr_name)
else:
print("Single-probe gene is probably CGH:", curr_name,
file=sys.stderr)
def exclude_regions(bed_fname, access_rows):
ex_by_chrom = dict(
group_regions_by_chromosome(parse_regions(bed_fname, coord_only=True)))
if len(ex_by_chrom) == 0:
# Nothing to exclude -> emit the input regions unmodified
for row in access_rows:
yield row
else:
# Check if each input region overlaps an excluded region
for chrom, a_rows in group_regions_by_chromosome(access_rows):
if chrom in ex_by_chrom:
logging.info("%s: Subtracting excluded regions", chrom)
exclude_rows = iter(ex_by_chrom[chrom])
ex_start, ex_end = next_or_inf(exclude_rows)
for a_start, a_end in a_rows:
for row in exclude_in_region(exclude_rows, chrom, a_start,
a_end, ex_start, ex_end):
yield row
else:
logging.info("%s: No excluded regions", chrom)
for a_start, a_end in a_rows:
yield (chrom, a_start, a_end)
def exclude_regions(bed_fname, access_rows):
ex_by_chrom = dict(group_regions_by_chromosome(
parse_regions(bed_fname, coord_only=True)))
if len(ex_by_chrom) == 0:
# Nothing to exclude -> emit the input regions unmodified
for row in access_rows:
yield row
else:
# Check if each input region overlaps an excluded region
for chrom, a_rows in group_regions_by_chromosome(access_rows):
if chrom in ex_by_chrom:
logging.info("%s: Subtracting excluded regions", chrom)
exclude_rows = iter(ex_by_chrom[chrom])
ex_start, ex_end = next_or_inf(exclude_rows)
for a_start, a_end in a_rows:
for row in exclude_in_region(exclude_rows, chrom, a_start,
a_end, ex_start, ex_end):
yield row
else:
logging.info("%s: No excluded regions", chrom)
for a_start, a_end in a_rows:
yield (chrom, a_start, a_end)
def interval2genes(interval, skip=('CGH', '-')):
"""Squash intervals into named genes."""
curr_name = None
curr_chrom = None
curr_start = None
curr_end = None
curr_len = 0
for chrom, start, end, name in ngfrills.parse_regions(interval):
if name in skip:
continue
if chrom != curr_chrom or name != curr_name:
if curr_name is not None:
# yield (curr_chrom, curr_start, curr_end, curr_name)
if curr_len > 1:
# Emit
yield (curr_chrom, curr_start, curr_end, curr_name)
else:
print("Single-probe gene is probably CGH:",
curr_name,
file=sys.stderr)
# Reset
curr_name = name
curr_chrom = chrom
curr_start = start
curr_len = 0
# Extend
curr_end = end
curr_len += 1
if curr_name is not None:
if curr_len > 1:
# Emit
yield (curr_chrom, curr_start, curr_end, curr_name)
else:
print("Single-probe gene is probably CGH:",
curr_name,
file=sys.stderr)