def test_complement(self):
genome_file = make_file_from_list(
[
['1', '2000'],
['2', '1000'],
['MT', '500'],
], bedtool=False)
genes = list_to_intervals([
['1', '.', 'gene1', '200', '400', '.', '+', '.', '.'],
['1', '.', 'gene2', '300', '600', '.', '+', '.', '.'],
['1', '.', 'gene3', '200', '500', '.', '+', '.', '.'],
['2', '.', 'gene4', '100', '200', '.', '+', '.', '.'],
['2', '.', 'gene5', '100', '300', '.', '-', '.', '.'],
])
complement = make_list_from_file(segment._complement(genes, genome_file, '+'), fields_separator='\t')
empty_col8 = 'ID "inter%s"; gene_id "."; transcript_id ".";'
expected = [
['1', '.', 'intergenic', '1', '199', '.', '+', '.', empty_col8 % "P00000"],
['1', '.', 'intergenic', '601', '2000', '.', '+', '.', empty_col8 % "P00001"],
['2', '.', 'intergenic', '1', '99', '.', '+', '.', empty_col8 % "P00002"],
['2', '.', 'intergenic', '201', '1000', '.', '+', '.', empty_col8 % "P00003"],
['MT', '.', 'intergenic', '1', '500', '.', '+', '.', empty_col8 % "P00004"],
]
self.assertEqual(complement, expected)
def make_types_length_file(annotation,
fai,
out_file=None,
subtype='biotype',
excluded_types=None):
"""
Calculate the number of non-overlapping base pairs of each "type".
In the context of this function "type" equals to the combination of 3rd
column and attribute subtype from annotation file (GTF).
Parameters
----------
annotation : str
Path to annotation GTF file (should include subtype attribute).
out_file : str
Path to output file (if None, name is set automatically).
subtype : int
Subtype.
excluded_types : list_str
Types listed in 3rd column of GTF to be exclude from analysis.
Returns
-------
str
Absolute path to out_file.
"""
excluded_types = excluded_types or []
ann_filtered = pybedtools.BedTool(annotation).filter(
lambda x: x[2] not in excluded_types).sort().saveas()
fai = _first_two_columns(fai)
if out_file is None:
match = re.match(r'([\w_]+\.\d+.).*', os.path.basename(annotation))
if match:
out_file = str(match.group(1)) + 'types_length.txt'
else:
out_file = annotation + 'types_length.txt'
# Merge all annotation and take what is left = unannotated regions:
unann_pos = _complement(ann_filtered.fn, fai, '+', type_name='unannotated')
unann_neg = _complement(ann_filtered.fn, fai, '-', type_name='unannotated')
ann_modified_file = out_file + 'modified'
ann_joined = ann_filtered.cat(unann_pos, unann_neg, postmerge=False). \
sort().saveas(ann_modified_file)
data = {}
for interval in ann_joined:
if subtype:
sbtyp = interval.attrs.get(subtype, None)
type_ = ' '.join([interval[2], sbtyp] if sbtyp else [interval[2]])
else:
type_ = interval[2]
if type_ not in data:
data[type_] = []
data[type_].append(create_interval_from_list(interval.fields))
type_lengths = {}
for type_, list_ in data.items():
total_bp = 0
# pylint: disable=unexpected-keyword-arg
for feature in pybedtools.BedTool(i for i in list_).merge(s=True):
total_bp += len(feature)
type_lengths[type_] = total_bp
# Write results to file:
with open(out_file, 'wt') as outfile:
for type_, length in sorted(type_lengths.items()):
outfile.write('{}\t{}\n'.format(type_, length))
return os.path.abspath(out_file), os.path.abspath(ann_modified_file)