def intron_sizes(
inputfile=None,
outputfile=None,
key_name=None):
"""
Add a new key to transcript features containing a comma-separated list of intron sizes.
"""
gtf = GTF(inputfile, check_ensembl_format=False)
all_tx_ids = gtf.get_tx_ids(nr=True)
intron_bo = gtf.get_introns(by_transcript=True,
name=["transcript_id"],
intron_nb_in_name=False,
feat_name=False)
strands = gtf.select_by_key("feature",
"transcript").extract_data("transcript_id,strand",
as_dict_of_values=True,
no_na=True,
nr=True,
hide_undef=True)
intron_size = {tx: [] for tx in all_tx_ids}
for bed_line in intron_bo:
intron_size[bed_line.name] += [str(bed_line.end - bed_line.start)]
for tx_id in intron_size:
if len(intron_size[tx_id]):
if strands[tx_id] == "-":
intron_size[tx_id] = ",".join(reversed(intron_size[tx_id]))
else:
intron_size[tx_id] = ",".join(intron_size[tx_id])
else:
intron_size[tx_id] = "0"
if len(intron_size):
gtf = gtf.add_attr_from_dict(feat="transcript",
key="transcript_id",
a_dict=intron_size,
new_key=key_name)
gtf.write(outputfile,
gc_off=True)
close_properly(outputfile, inputfile)
def exon_sizes(inputfile=None, outputfile=None, key_name=None):
"""
Add a new key to transcript features containing a comma-separated list of exon-size.
"""
gtf = GTF(inputfile)
all_tx_ids = gtf.get_tx_ids(nr=True)
tx_to_size_list = dict()
exons_starts = gtf.select_by_key("feature", "exon").extract_data(
"transcript_id,start",
as_dict_of_merged_list=True,
no_na=True,
nr=False)
if not len(exons_starts):
message("No exon found.", type="ERROR")
exons_ends = gtf.select_by_key("feature", "exon").extract_data(
"transcript_id,end", as_dict_of_merged_list=True, no_na=True, nr=False)
strands = gtf.select_by_key("feature", "transcript").extract_data(
"transcript_id,strand",
as_dict_of_values=True,
no_na=True,
nr=True,
hide_undef=True)
for tx_id in all_tx_ids:
size_list = []
for s, e in zip(exons_starts[tx_id], exons_ends[tx_id]):
size = str(int(e) - int(s) + 1)
size_list += [size]
if strands[tx_id] == "-":
size_list = reversed(size_list)
tx_to_size_list[tx_id] = ",".join(size_list)
if len(tx_to_size_list):
gtf = gtf.add_attr_from_dict(feat="transcript",
key="transcript_id",
a_dict=tx_to_size_list,
new_key=key_name)
gtf.write(outputfile, gc_off=True)
close_properly(outputfile, inputfile)
def select_by_intron_size(inputfile=None,
outputfile=None,
intron_size=0,
merged=False,
invert_match=False,
delete_monoexonic=False,
add_intron_size=False):
"""
Select genes which contain an intron of size at least s or whose sum of intron size is at least s
"""
message("Searching for intronic regions.")
gtf = GTF(inputfile, check_ensembl_format=False)
introns_bo = gtf.get_introns(by_transcript=True,
name=["transcript_id"],
intron_nb_in_name=False).sort()
# Get the list of transcripts
all_tx_ids = gtf.get_tx_ids(nr=True)
# The list of transcripts
# to be deleted
to_delete = OrderedDict()
if merged:
# Create a dict that will contain the sum of introns for
# each transcript
intron_sum_dict = OrderedDict.fromkeys(all_tx_ids, 0)
for i in introns_bo:
size = i.end - i.start
tx_id = i.name
intron_sum_dict[tx_id] += size
for tx_id, sum_intron in list(intron_sum_dict.items()):
if sum_intron != 0:
if not invert_match:
if sum_intron < intron_size:
to_delete[tx_id] = 1
else:
if sum_intron >= intron_size:
to_delete[tx_id] = 1
else:
if delete_monoexonic:
to_delete[tx_id] = 1
if add_intron_size:
gtf = gtf.add_attr_from_dict(feat="transcript",
key="transcript_id",
a_dict=intron_sum_dict,
new_key="intron_size_sum")
else:
# Create a dict that will contain a list introns size
# for each transcript
intron_size_dict = defaultdict(list)
for tx_id in all_tx_ids:
intron_size_dict[tx_id] = []
for i in introns_bo:
size = i.end - i.start
tx_id = i.name
intron_size_dict[tx_id] += [size]
for tx_id, list_size in list(intron_size_dict.items()):
if not list_size:
intron_size_dict[tx_id] = [0]
if delete_monoexonic:
to_delete[tx_id] = 1
continue
for size in intron_size_dict[tx_id]:
if not invert_match:
if size < intron_size:
to_delete[tx_id] = 1
else:
if size >= intron_size:
to_delete[tx_id] = 1
if add_intron_size:
for tx_id, list_size in list(intron_size_dict.items()):
list_size = [str(x) for x in list_size]
intron_size_dict[tx_id] = ",".join(list_size)
gtf = gtf.add_attr_from_dict(feat="transcript",
key="transcript_id",
a_dict=intron_size_dict,
new_key="intron_size")
all_tx_ids = gtf.get_tx_ids(nr=True)
all_tx_ids = [x for x in all_tx_ids if x not in to_delete]
msg_list = ",".join(list(to_delete.keys()))
nb_char = min([len(msg_list), 40])
msg_list = msg_list[0:nb_char]
message("Deleting: " + msg_list + "...")
gtf = gtf.select_by_key("transcript_id", ",".join(all_tx_ids))
gtf.write(outputfile, gc_off=True)
close_properly(outputfile, inputfile)
