def count_feats(gbkf, verbose=False):
if verbose:
message(f"Reading {gbkf}", "BLUE")
count = {}
for seq in genbank_seqio(gbkf):
for feat in seq.features:
count[feat.type] = count.get(feat.type, 0) + 1
return count
def hmmsearch_print_then_parse(gbkf, hmmf):
aaout = NamedTemporaryFile(mode='w+t', delete=False)
sys.stderr.write(f"Writing the amino acids to {aaout.name}\n")
aaout.seek(0)
for seq in genbank_seqio(gbkf):
for feat in seq.features:
if feat.type != 'CDS':
continue
aa = ""
if 'translation' in feat.qualifiers:
aa = feat.qualifiers['translation'][0]
else:
aa = str(feat.extract(seq).translate().seq)
myid = feature_id(seq, feat)
aaout.write(f">{myid}\n{aa}\n")
aaout.close()
sys.stderr.write("Searching\n")
try:
search = subprocess.Popen([
"hmmsearch", '--cpu', '6', '-E', '1e-10', '--domE', '1e-5',
'--noali', hmmf, aaout.name
],
stdout=subprocess.PIPE)
except subprocess.CalledProcessError as e:
sys.stderr.write(f"Error running hmmscan:\n{e}\n")
sys.exit(-1)
sys.stderr.write("Parsing\n")
hmmresult = search.communicate()[0]
results = SearchIO.parse(StringIO(hmmresult.decode()), 'hmmer3-text')
allhits = {}
hitcount = 0
rescount = 0
for res in results:
allhits[res.id] = {}
rescount += 1
for hit in res:
allhits[res.id][hit.id] = hit.evalue
# print(f"Result: {res.id}: Hit: {hit.id} Eval: {hit.evalue}")
hitcount += 1
print(
f"Using hmmsearch and tempfiles there were {rescount} results and {hitcount} hits, and our dict has {len(allhits)} entries"
)
def stream_hmmsearch(gbkf, hmmf):
"""
NOTE: THIS DOES NOT WORK!!
You can't stream against hmmsearch as it can't rewind the sequences. You either need to use hmmscan (slow) or a temp file
(fast!)
:param gbkf:
:param hmmf:
:return:
"""
allhits = {}
hitcount = 0
rescount = 0
for seq in genbank_seqio(gbkf):
prots = []
for feat in seq.features:
if feat.type != 'CDS':
continue
aa = ""
if 'translation' in feat.qualifiers:
aa = feat.qualifiers['translation'][0]
else:
aa = str(feat.extract(seq).translate().seq)
myid = feature_id(seq, feat)
prots.append(f">{myid}\n{aa}")
search = subprocess.Popen([
"hmmsearch", '--cpu', '6', '-E', '1e-10', '--domE', '1e-5',
'--noali', hmmf, '-'
],
stdin=subprocess.PIPE,
stdout=subprocess.PIPE)
hmmresult = search.communicate(input="\n".join(prots).encode())[0]
results = SearchIO.parse(StringIO(hmmresult.decode()), 'hmmer3-text')
for res in results:
allhits[res.id] = {}
rescount += 1
for hit in res:
allhits[res.id][hit.id] = hit.evalue
hitcount += 1
print(
f"Using hmmsearch and streaming all at once there were {rescount} results and {hitcount} hits, and our dict has {len(allhits)} entries"
)
def run_hmmscan_aao(gbkf, hmmf):
allhits = {}
hitcount = 0
rescount = 0
for seq in genbank_seqio(gbkf):
prots = []
for feat in seq.features:
if feat.type != 'CDS':
continue
aa = ""
if 'translation' in feat.qualifiers:
aa = feat.qualifiers['translation'][0]
else:
aa = str(feat.extract(seq).translate().seq)
myid = feature_id(seq, feat)
prots.append(f">{myid}\n{aa}")
search = subprocess.Popen([
"hmmscan", '--cpu', '6', '-E', '1e-10', '--domE', '1e-5',
'--noali', hmmf, '-'
],
stdin=subprocess.PIPE,
stdout=subprocess.PIPE)
hmmresult = search.communicate(input="\n".join(prots).encode())[0]
results = SearchIO.parse(StringIO(hmmresult.decode()), 'hmmer3-text')
for res in results:
allhits[res.id] = {}
rescount += 1
for hit in res:
allhits[res.id][hit.id] = hit.evalue
hitcount += 1
print(
f"Using hmmscan and streaming all at once there were {rescount} results and {hitcount} hits, and our dict has {len(allhits)} entries"
)
Test a directory of genbank files and note whether they have the is_phage qualifier for their genomes
"""
import os
import sys
import argparse
from roblib import genbank_seqio, message
__author__ = 'Rob Edwards'
__copyright__ = 'Copyright 2020, Rob Edwards'
__credits__ = ['Rob Edwards']
__license__ = 'MIT'
__maintainer__ = 'Rob Edwards'
__email__ = '*****@*****.**'
if __name__ == '__main__':
parser = argparse.ArgumentParser(description=" ")
parser.add_argument('-d', help='directory of genbank files', required=True)
parser.add_argument('-v', help='verbose output', action='store_true')
args = parser.parse_args()
for f in os.listdir(args.d):
if args.v:
message(f"Reading {f}", "GREEN")
pc = 0
for s in genbank_seqio(os.path.join(args.d, f)):
for feat in s.features:
if 'is_phage' in feat.qualifiers:
pc += 1
print(f"{f}\t{pc}")