def genePairTAD(genes, tads, fraction=0.7):
"""
Annotate tads with genes and return as dict.
Params:
--------
genes: `str` bed4 file of gene
tads: `str` bed3 file of tad
fraction: `str` or `float` fraction of gene
overlap with tads [default: 0.7]
Returns:
--------
out: `dict` dictionary of TADs annotation
Examples:
--------
>>> db = TADConserved().getGene("gene.bed", "tad.bed")
"""
check_file_exists(tads)
check_file_exists(genes)
if 0 > float(fraction) > 1:
logging.error('The option `-f` must set in '
'range [0, 1], and you set {}'.format(fraction))
sys.exit()
bedtools_cmd = "bedtools intersect -a {} -b {} -wao -f {} | cut -f 4-7 ".format(
genes, tads, fraction)
db = OrderedDict()
for line in os.popen(bedtools_cmd):
line_list = line.strip().split()
gene, chrom, start, end = line_list
ID = chrRangeID([chrom, start, end]) if chrom != "." \
else "."
if ID == ".":
continue
db[gene] = ID
return db
def testPipe(args):
"""
test pipe for tad annotate.
"""
db = OrderedDict()
if not sys.stdin.isatty():
handle = sys.stdin
else:
pass
for line in sys.stdin:
line_list = line.strip().split()
ID = chrRangeID(line_list[:3])
gene = line_list[3]
if ID not in db:
db[ID] = []
db[ID].append(gene)
for ID in db:
print(ID + "\t" + ",".join(db[ID]))
def annotate(args):
"""
%prog tad.bed gene.bed [Options]
Annotate tads with gene.
"""
p = OptionParser(annotate.__doc__)
p.add_option('-F',
dest='fraction',
default='0.7',
help='the fraction of gene overlap of tads'
' [default: %default]')
p.add_option('--isnum',
default=False,
action='store_true',
help='if output the gene number [default: %default]')
p.add_option('--plot',
default=False,
action='store_true',
help='if plot the gene number '
'distribution [default: %default]')
opts, args = p.parse_args(args)
if len(args) != 2:
sys.exit(p.print_help())
tads, genes = args
fraction = opts.fraction
db = TADConserved().getGene(tads, genes, fraction, opts.isnum, opts.plot)
if opts.isnum:
return
for ID in db:
gene_list = sorted(db[ID])
length = len(gene_list) if "." not in gene_list else 0
print("\t".join(chrRangeID(ID, axis=1)) + "\t" + \
",".join(gene_list) + "\t" + \
str(length), file=sys.stdout)
def getGene(tads, genes, fraction=0.7, isnum=False, isPlot=False):
"""
Annotate tads with genes and return as dict.
Params:
--------
tads: `str` bed3 file of tad
genes: `str` bed4 file of gene
fraction: `str` or `float` fraction of gene
overlap with tads [default: 0.7]
isnum: `bool` if set output the gene number instead
of gene list. [default: False]
isPlot: `bool` if plot the gene number per TADs
distribution. [default: False]
Returns:
--------
out: `dict` dictionary of TADs annotation
Examples:
--------
>>> db = TADConserved().getGene("tad.bed", "gene.bed")
"""
check_file_exists(tads)
check_file_exists(genes)
if 0 > float(fraction) > 1:
logging.error('The option `-F` must set in '
'range [0, 1], and you set {}'.format(fraction))
sys.exit()
bedtools_cmd = "bedtools intersect -a {} -b {} -wao -F {} | \
cut -f 1-3,7 ".format(tads, genes, fraction)
db = OrderedDict()
for line in os.popen(bedtools_cmd):
line_list = line.strip().split()
ID = chrRangeID(line_list[:3])
gene = line_list[3]
if ID not in db:
db[ID] = set()
db[ID].add(gene)
if isnum:
for ID in db:
db[ID] = len(db[ID])
if isPlot:
assert isnum, 'isnum must specify as True'
fig, ax = plt.subplots(figsize=(5, 5))
sns.distplot(db.values(), hist=False, kde=True, ax=ax)
ax.set_xticks(range(0, 41, 5))
ax.set_xlim(0, 40)
ax.set_xlabel('Gene number')
ax.set_ylabel('Frequence')
ax.set_title('Gene Number Distribution ({:,})'.format(
sum(db.values())))
plt.savefig('{}.gene_num_dist.pdf'.format(genes.rsplit('.', 1)[0]),
dpi=300)
logging.debug('Successful to plot gene number distribution '
'`{}.gene_num_dist.pdf`.'.format(
genes.rsplit('.', 1)[0]))
return db