def runExonerate(input):
s = input.split(':::')
ProtID = s[0]
ScaffID = s[1]
ScaffStart = int(s[2])
ScaffEnd = int(s[3])
#get the protein model
query = os.path.join(tmpdir, ProtID + '.' + str(os.getpid()) + '.fa')
with open(query, 'w') as output:
SeqIO.write(protein_dict[ProtID], output, 'fasta')
#now get the genome region, use different variable names for SeqRecords to avoid collision
scaffold = os.path.join(
tmpdir, ScaffID + '.' + ProtID + '.' + str(ScaffStart) + '-' +
str(ScaffEnd) + '.fa')
with open(scaffold, 'w') as output2:
with open(os.path.join(tmpdir, 'scaffolds', ScaffID + '.fa'),
'rU') as fullscaff:
for header, Sequence in SimpleFastaParser(fullscaff):
#grab a 3 kb cushion on either side of hit region, careful of scaffold ends
start = ScaffStart - 3000
if start < 1:
start = 1
end = ScaffEnd + 3000
if end > len(Sequence):
end = len(Sequence)
output2.write('>%s\n%s\n' % (header, Sequence[start:end]))
exoname = ProtID + '.' + ScaffID + '__' + str(start) + '__'
#check that input files are created and valid
exonerate_out = os.path.join(tmpdir, 'exonerate.' + exoname + '.out')
ryo = "AveragePercentIdentity: %pi\n"
cmd = [
'exonerate', '--model', 'p2g', '--showvulgar', 'no', '--showalignment',
'no', '--showquerygff', 'no', '--showtargetgff', 'yes', '--maxintron',
str(args.maxintron), '--percent', '80', '--ryo', ryo, query, scaffold
]
if lib.checkannotations(query) and lib.checkannotations(scaffold):
#run exonerate, capture errors
with open(exonerate_out, 'w') as output3:
proc = subprocess.Popen(cmd,
stdout=output3,
stderr=subprocess.PIPE)
stderr = proc.communicate()
if 'WARNING' in stderr[1]:
lib.log.debug('Error in input:{:}'.format(input))
lib.log.debug(
'%s, Len=%i, %i-%i; %i-%i' %
(header, len(Sequence), ScaffStart, ScaffEnd, start, end))
os.rename(query,
os.path.join(tmpdir, 'failed', os.path.basename(query)))
os.rename(
scaffold,
os.path.join(tmpdir, 'failed', os.path.basename(scaffold)))
else:
for y in [query, scaffold]:
try:
lib.SafeRemove(y)
except OSError:
lib.log.debug("Error removing %s" % (y))
#check filesize of exonerate output, no hits still have some output data in them, should be safe dropping anything smaller than 500 bytes
if lib.getSize(exonerate_out) < 500:
os.remove(exonerate_out)
else:
lib.log.debug('Error in query or scaffold:{:}'.format(input))
lib.SafeRemove(query)
lib.SafeRemove(scaffold)
output.write('\t'.join(cols))
#convert to GFF3 using ExoConverter from EVM
with open(args.out, 'w') as output:
subprocess.call([ExoConverter, exonerate_raw], stdout=output, stderr=FNULL)
#output some quick summary of exonerate alignments that you found
Found = lib.countGFFgenes(exonerate_raw)
lib.log.info('Exonerate finished: found {:,} alignments'.format(Found))
#check for saving output of tblastn
if args.tblastn_out:
shutil.copyfile(BlastResult, args.tblastn_out)
#finally clean-up your mess if failed is empty
if args.debug:
try:
os.rmdir(os.path.join(tmpdir, 'failed'))
empty = True
except OSError:
empty = False
if empty:
lib.SafeRemove(tmpdir)
else:
lib.log.error("Failed exonerate alignments found, see files in %s" %
os.path.join(tmpdir, 'failed'))
else:
if os.path.isdir(tmpdir):
lib.SafeRemove(tmpdir)
sys.exit(1)
#create tmp folder to run tbl2asn from
#make tmp folder
tmp = outputname + '_tmp'
if not os.path.exists(tmp):
os.makedirs(tmp)
#now move files into proper location
if not lib.checkannotations(args.fasta):
print('FASTA genome file not found: {:}'.format(args.fasta))
sys.exit(1)
if not lib.checkannotations(args.tbl):
print('TBL annotations file not found: {:}'.format(args.tbl))
sys.exit(1)
shutil.copyfile(args.fasta, os.path.join(tmp, 'genome.fsa'))
shutil.copyfile(args.tbl, os.path.join(tmp, 'genome.tbl'))
#now we can run tbl2asn
if args.sbt:
SBT = args.sbt
else:
SBT = os.path.join(parentdir, 'lib', 'test.sbt')
discrep = outputname + '.discrepency.txt'
version = 1
tbl2asn_cmd = runtbl2asn(tmp, SBT, discrep, organism, args.isolate,
args.strain, args.tbl2asn, version)
#get output files
gbkout = outputname + '.gbk'
shutil.copyfile(os.path.join(tmp, 'genome.gbf'), gbkout)
lib.SafeRemove(tmp)