def parse_blastxml(input_path, augustus_mapping, feature_table,
annotation_count_with_putative_function, gene_counter,
locus_tag, min_coverage, min_ident):
#print input_path, augustus_mapping, feature_table, annotation_count_with_putative_function, gene_counter, locus_tag
# extract the sequence name
seq_name = os.path.splitext(os.path.split(input_path)[-1])[0]
#input_path.split('Seq')[-1].split('.blastxml')[0]
#print seq_name
feature_table.write('>Feature %s\n' % seq_name)
#locus_tag = 'M7I_'
with open(input_path) as blast_handle:
for entry in NCBIXML.parse(blast_handle):
if entry.application == "BLASTX":
query_length = entry.query_length / 3
if type(query_length) == type(1.7):
print "Query length is not a multiple of three"
break
query_id = entry.query.split()[0]
query_info = augustus_mapping[query_id]
assert query_info.mRNA.seq_type == 'gene'
gene_start = query_info.mRNA.start
gene_end = query_info.mRNA.stop
cds = query_info.exons
mRNA = query_info.mRNA
else:
break
gene_counter += 1
hsp_has_annotation = False
feature_table_text = dict()
for alignment in entry.alignments:
for hsp in alignment.hsps:
nident = hsp.identities
ident = (100 * float(nident) / float(hsp.align_length))
"""
Coverage: 'c8-c7+1 >= 0.5*c23'
"""
coverage = False
if int(hsp.query_end) - int(
hsp.query_start
) + 1 >= min_coverage * query_length:
coverage = True
# only annotate hits with an identity over 50% and a coverage over 50%
if ident > min_ident and coverage:
feature_table_text[hsp.bits] = ""
hsp_has_annotation = True
"""
Hit_def changed: It now looks like:
'RecName: Full=Erythronolide synthase, modules 3 and 4; Short=PKS; AltName: Full=6-deoxyerythronolide B synthase II; AltName: Full=DEBS 2; AltName: Full=ORF 2'
"""
print alignment.hit_def
accession = alignment.hit_def.encode('utf8')
accession = filter(
lambda token: token.startswith('RecName:'),
map(str.strip,
accession.split(';')))[0].split('Full=')[-1]
accession = change_according_reviewer(accession,
note_line=False)
feature_table_text[hsp.bits] += '%i\t%i\tgene\n' % (
gene_start, gene_end)
feature_table_text[
hsp.bits] += '\t\t\tlocus_tag\t%s%04d\n' % (
locus_tag, gene_counter)
short_intron = check_short_introns(cds)
if short_intron:
feature_table_text[hsp.bits] += '\t\t\tpseudo\n'
feature_table_text[
hsp.
bits] += '\t\t\tnote\tnonfunctional; similar to %s\n' % accession
continue
"""
Write the CDS section for the 'annotation' case and save a string for the mRNA section.
"""
mRNA_annotation = ''
mRNA_annotation += '%i\t%i\tmRNA\n' % (cds[0].start,
cds[0].stop)
feature_table_text[hsp.bits] += '%i\t%i\tCDS\n' % (
cds[0].start, cds[0].stop)
for region in cds[1:]:
feature_table_text[hsp.bits] += '%i\t%i\n' % (
region.start, region.stop)
mRNA_annotation += '%i\t%i\n' % (region.start,
region.stop)
if accession.startswith('hypothetical protein') or \
accession.startswith('predicted protein') or \
accession == '' or accession == 'protein':
feature_table_text[
hsp.
bits] += '\t\t\tproduct\thypothetical protein\n'
mRNA_annotation += '\t\t\tproduct\thypothetical protein\n'
else:
feature_table_text[
hsp.bits] += '\t\t\tproduct\tputative %s\n' % (
accession)
mRNA_annotation += '\t\t\tproduct\tputative %s\n' % (
accession)
feature_table_text[
hsp.
bits] += '\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' % (
locus_tag, gene_counter)
mRNA_annotation += '\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' % (
locus_tag, gene_counter)
feature_table_text[
hsp.
bits] += '\t\t\ttranscript_id\tgnl|PBUF|%smrna%04d\n' % (
locus_tag, gene_counter)
mRNA_annotation += '\t\t\ttranscript_id\tgnl|PBUF|%smrna%04d\n' % (
locus_tag, gene_counter)
# Write mRNA section
feature_table_text[hsp.bits] += mRNA_annotation
if str(hsp.expect).find('e') != -1:
""" Der evalue ist eine lange Zahl und muss gekuertzt werden. Z.B. 4.787347812347e-124"""
evalue_first, evalue_last = str(
hsp.expect).split('e')
evalue = str(round(float(evalue_first),
1)) + 'e' + evalue_last
else:
evalue = round(hsp.expect, 1)
"""
hit_def = change_according_reviewer(alignment.hit_def, note_line = True)
if hit_def.split('|')[:-1] != []:
hit_def = hit_def.split('|')[-1].split()[0]
else:
hit_def = accession
"""
hit_def = accession
"""
try:
protein_accession_gb = hit_def.split('gb|')[1].split('|')[0] #try to extract the genbank accession number >gi|302432474|gb|EFL04290.1|; -> EFL04290.1
inference = "similar to AA sequence:INSD: %s" % protein_accession_gb
feature_table_text[ hsp.bits ] += '\t\t\tinference\t%s\n' % (inference)
protein_accession_ref = hit_def.split('ref|')[1].split('|')[0] #try to extract the genbank accession number >gi|302432474|gb|EFL04290.1|; -> EFL04290.1
inference = "similar to AA sequence:RefSeq: %s" % protein_accession_ref
feature_table_text[ hsp.bits ] += '\t\t\tinference\t%s\n' % (inference)
except:
pass
"""
inference = """ab initio prediction:Augustus:2.5.5"""
feature_table_text[
hsp.bits] += '\t\t\tinference\t%s\n' % (inference)
note = """similar to UniProtKB/Swiss-Prot Entry: %(hit_accession)s""" % {
'gene_counter': gene_counter,
'accession': accession,
'alignment_hit_def': hit_def,
'hit_accession': alignment.accession,
'len': query_length,
'evalue': evalue,
'bit_score': round(hsp.bits, 2),
'locus_tag': locus_tag,
}
feature_table_text[
hsp.bits] += '\t\t\tnote\t%s\n' % (note)
#for region in cds[1:]:
# mRNA_annotation += '%i\t%i\n' % (region.start, region.stop)
# feature_table.write('%i\t%i\n' % (region.start, region.stop))
if hsp_has_annotation == False:
"""
If hsp has no annotation, insert a hypothetical protein
"""
feature_table.write('%i\t%i\tgene\n' % (gene_start, gene_end))
feature_table.write('\t\t\tlocus_tag\t%s%04d\n' %
(locus_tag, gene_counter))
assert cds[0].seq_type == 'CDS'
short_intron = check_short_introns(cds)
if short_intron:
feature_table.write('\t\t\tpseudo\n')
feature_table.write('\t\t\tnote\tnonfunctional\n')
"""
Write the CDS section for the 'no-annotation' case.
"""
feature_table.write('%i\t%i\tCDS\n' %
(cds[0].start, cds[0].stop))
for region in cds[1:]:
feature_table.write('%i\t%i\n' %
(region.start, region.stop))
feature_table.write('\t\t\tproduct\thypothetical protein\n')
feature_table.write('\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' %
(locus_tag, gene_counter))
feature_table.write(
'\t\t\ttranscript_id\tgnl|PBUF|%smrna%04d\n' %
(locus_tag, gene_counter))
feature_table.write(
'\t\t\tnote\tpredicted with Augustus 2.5.5\n')
"""
Write the mRNA section for the 'no-annotation' case.
"""
feature_table.write('%i\t%i\tmRNA\n' %
(cds[0].start, cds[0].stop))
for region in cds[1:]:
feature_table.write('%i\t%i\n' %
(region.start, region.stop))
feature_table.write('\t\t\tproduct\thypothetical protein\n')
feature_table.write('\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' %
(locus_tag, gene_counter))
feature_table.write(
'\t\t\ttranscript_id\tgnl|PBUF|%smrna%04d\n' %
(locus_tag, gene_counter))
else:
bitscores = feature_table_text.keys()
bitscores.sort(reverse=True)
feature_table.write(feature_table_text[bitscores[0]])
if feature_table_text[bitscores[0]].find(
'\t\t\tproduct\thypothetical protein\n') == -1:
annotation_count_with_putative_function += 1
return (gene_counter, annotation_count_with_putative_function)
def parse_blastxml(input_path, glimmer_mapping, feature_table,
annotation_count_with_putative_function, gene_counter,
locus_tag, min_coverage, min_ident):
# extract the sequence number
seq_number = input_path.split('Seq')[-1].split('.blastxml')[0]
feature_table.write('>Feature Seq%s\n' % seq_number)
with open(input_path) as blast_handle:
for entry in NCBIXML.parse(blast_handle):
if entry.application == "BLASTX":
query_length = entry.query_length
if type(query_length) == type(1.7):
print "Query length is not a multiple of three"
break
query_id = entry.query.split()[0]
query_info = glimmer_mapping[query_id]
query_start = int(query_info[0])
query_end = int(query_info[1])
else:
break
gene_counter += 1
"""
if not entry.alignments:
feature_table.write('%i\t%i\tgene\n' % (query_start, query_end))
feature_table.write('\t\t\tlocus_tag\t%s%04d\n' % (locus_tag, gene_counter))
feature_table.write('%i\t%i\tCDS\n' % (query_start, query_end))
feature_table.write('\t\t\tproduct\thypothetical protein\n')
feature_table.write('\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' % (locus_tag, gene_counter))
feature_table.write('\t\t\tnote\tpredicted with glimmer3\n')
break
"""
hsp_has_annotation = False
feature_table_text = dict()
for alignment in entry.alignments:
for hsp in alignment.hsps:
nident = hsp.identities
ident = (100 * float(nident) / float(hsp.align_length))
coverage = False
if int(hsp.query_end) - int(
hsp.query_start
) + 1 >= min_coverage * query_length:
coverage = True
# only annotate hits with an identity over 50% and a coverage over 50%
if ident > min_ident and coverage:
feature_table_text[hsp.bits] = ""
hsp_has_annotation = True
"""
Hit_def changed: It now looks like:
'RecName: Full=Erythronolide synthase, modules 3 and 4; Short=PKS; AltName: Full=6-deoxyerythronolide B synthase II; AltName: Full=DEBS 2; AltName: Full=ORF 2'
"""
print alignment.hit_def
accession = alignment.hit_def.encode('utf8')
accession = filter(
lambda token: token.startswith('RecName:'),
map(str.strip,
accession.split(';')))[0].split('Full=')[-1]
assert change_according_reviewer(
'Pimelyl-[acyl-carrier protein] methyl ester esterase',
note_line=False
) == 'Pimelyl-[acyl-carrier protein] methyl ester esterase'
assert change_according_reviewer(
'putative D-malate dehydrogenase [decarboxylating] [gnl|PBUF|STVIR_0046:1-352] [gnl|PBUF|STVIR_0046: raw, aa len= 352]',
note_line=False) == 'D-malate dehydrogenase'
accession = change_according_reviewer(accession,
note_line=False)
feature_table_text[hsp.bits] += '%i\t%i\tgene\n' % (
query_start, query_end)
feature_table_text[
hsp.bits] += '\t\t\tlocus_tag\t%s%04d\n' % (
locus_tag, gene_counter)
feature_table_text[hsp.bits] += '%i\t%i\tCDS\n' % (
query_start, query_end)
if accession.startswith('hypothetical protein') or \
accession.startswith('predicted protein') or \
accession == '' or accession == 'protein':
feature_table_text[
hsp.
bits] += '\t\t\tproduct\thypothetical protein\n'
else:
feature_table_text[
hsp.bits] += '\t\t\tproduct\tputative %s\n' % (
accession)
feature_table_text[
hsp.
bits] += '\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' % (
locus_tag, gene_counter)
if str(hsp.expect).find('e') != -1:
""" Der evalue ist eine lange Zahl und muss gekuertzt werden. Z.B. 4.787347812347e-124"""
evalue_first, evalue_last = str(
hsp.expect).split('e')
evalue = str(round(float(evalue_first),
1)) + 'e' + evalue_last
else:
evalue = round(hsp.expect, 1)
"""
hit_def = change_according_reviewer(alignment.hit_def, note_line = True)
if hit_def.split('|')[:-1] != []:
hit_def = hit_def.split('|')[-1].split()[0]
else:
hit_def = accession
"""
""""
try:
protein_accession_gb = hit_def.split('gb|')[1].split('|')[0] #try to extract the genbank accession number >gi|302432474|gb|EFL04290.1|; -> EFL04290.1
inference = "similar to AA sequence:INSD: %s" % protein_accession_gb
feature_table_text[ hsp.bits ] += '\t\t\tinference\t%s\n' % (inference)
protein_accession_ref = hit_def.split('ref|')[1].split('|')[0] #try to extract the genbank accession number >gi|302432474|gb|EFL04290.1|; -> EFL04290.1
inference = "similar to AA sequence:RefSeq: %s" % protein_accession_ref
feature_table_text[ hsp.bits ] += '\t\t\tinference\t%s\n' % (inference)
except:
pass
"""
inference = """ab initio prediction:Glimmer:3"""
feature_table_text[
hsp.bits] += '\t\t\tinference\t%s\n' % (inference)
note = """similar to UniProtKB/Swiss-Prot Entry: %(hit_accession)s""" % {
'gene_counter': gene_counter,
'accession': accession,
'alignment_hit_def': accession,
'hit_accession': alignment.accession,
'len': query_length,
'evalue': evalue,
'bit_score': round(hsp.bits, 2),
'locus_tag': locus_tag,
}
feature_table_text[
hsp.bits] += '\t\t\tnote\t%s\n' % (note)
if hsp_has_annotation == False:
"""
If hsp has no annotation with the specified identity and coverage, insert a hypothetical protein
"""
feature_table.write('%i\t%i\tgene\n' %
(query_start, query_end))
feature_table.write('\t\t\tlocus_tag\t%s%04d\n' %
(locus_tag, gene_counter))
feature_table.write('%i\t%i\tCDS\n' % (query_start, query_end))
feature_table.write('\t\t\tproduct\thypothetical protein\n')
feature_table.write('\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' %
(locus_tag, gene_counter))
feature_table.write(
'\t\t\tnote\tab initio prediction:Glimmer3\n')
else:
bitscores = feature_table_text.keys()
bitscores.sort(reverse=True)
feature_table.write(feature_table_text[bitscores[0]])
if feature_table_text[bitscores[0]].find(
'\t\t\tproduct\thypothetical protein\n') == -1:
annotation_count_with_putative_function += 1
return (gene_counter, annotation_count_with_putative_function)
def parse_blastxml(input_path, augustus_mapping, feature_table, annotation_count_with_putative_function, gene_counter, locus_tag, min_coverage, min_ident):
#print input_path, augustus_mapping, feature_table, annotation_count_with_putative_function, gene_counter, locus_tag
# extract the sequence name
seq_name = os.path.splitext(os.path.split(input_path)[-1])[0]
#input_path.split('Seq')[-1].split('.blastxml')[0]
#print seq_name
feature_table.write('>Feature %s\n' % seq_name)
#locus_tag = 'M7I_'
with open(input_path) as blast_handle:
for entry in NCBIXML.parse(blast_handle):
if entry.application == "BLASTX":
query_length = entry.query_length / 3
if type(query_length) == type(1.7):
print "Query length is not a multiple of three"
break
query_id = entry.query.split()[0]
query_info = augustus_mapping[ query_id ]
assert query_info.mRNA.seq_type == 'gene'
gene_start = query_info.mRNA.start
gene_end = query_info.mRNA.stop
cds = query_info.exons
mRNA = query_info.mRNA
else:
break
gene_counter += 1
hsp_has_annotation = False
feature_table_text = dict()
for alignment in entry.alignments:
for hsp in alignment.hsps:
nident = hsp.identities
ident = (100*float(nident)/float(hsp.align_length))
"""
Coverage: 'c8-c7+1 >= 0.5*c23'
"""
coverage = False
if int(hsp.query_end) - int(hsp.query_start) + 1 >= min_coverage * query_length:
coverage = True
# only annotate hits with an identity over 50% and a coverage over 50%
if ident > min_ident and coverage:
feature_table_text[ hsp.bits ] = ""
hsp_has_annotation = True
"""
Hit_def changed: It now looks like:
'RecName: Full=Erythronolide synthase, modules 3 and 4; Short=PKS; AltName: Full=6-deoxyerythronolide B synthase II; AltName: Full=DEBS 2; AltName: Full=ORF 2'
"""
print alignment.hit_def
accession = alignment.hit_def.encode('utf8')
accession = filter(lambda token: token.startswith('RecName:'), map(str.strip, accession.split(';')))[0].split('Full=')[-1]
accession = change_according_reviewer(accession, note_line = False)
feature_table_text[ hsp.bits ] += '%i\t%i\tgene\n' % (gene_start, gene_end)
feature_table_text[ hsp.bits ] += '\t\t\tlocus_tag\t%s%04d\n' % (locus_tag, gene_counter)
short_intron = check_short_introns(cds)
if short_intron:
feature_table_text[ hsp.bits ] += '\t\t\tpseudo\n'
feature_table_text[ hsp.bits ] += '\t\t\tnote\tnonfunctional; similar to %s\n' % accession
continue
"""
Write the CDS section for the 'annotation' case and save a string for the mRNA section.
"""
mRNA_annotation = ''
mRNA_annotation += '%i\t%i\tmRNA\n' % (cds[0].start, cds[0].stop)
feature_table_text[ hsp.bits ] += '%i\t%i\tCDS\n' % (cds[0].start, cds[0].stop)
for region in cds[1:]:
feature_table_text[ hsp.bits ] += '%i\t%i\n' % (region.start, region.stop)
mRNA_annotation += '%i\t%i\n' % (region.start, region.stop)
if accession.startswith('hypothetical protein') or \
accession.startswith('predicted protein') or \
accession == '' or accession == 'protein':
feature_table_text[ hsp.bits ] += '\t\t\tproduct\thypothetical protein\n'
mRNA_annotation += '\t\t\tproduct\thypothetical protein\n'
else:
feature_table_text[ hsp.bits ] += '\t\t\tproduct\tputative %s\n' % (accession)
mRNA_annotation += '\t\t\tproduct\tputative %s\n' % (accession)
feature_table_text[ hsp.bits ] += '\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' % (locus_tag, gene_counter)
mRNA_annotation += '\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' % (locus_tag, gene_counter)
feature_table_text[ hsp.bits ] += '\t\t\ttranscript_id\tgnl|PBUF|%smrna%04d\n' % (locus_tag, gene_counter)
mRNA_annotation += '\t\t\ttranscript_id\tgnl|PBUF|%smrna%04d\n' % (locus_tag, gene_counter)
# Write mRNA section
feature_table_text[ hsp.bits ] += mRNA_annotation
if str(hsp.expect).find('e') != -1:
""" Der evalue ist eine lange Zahl und muss gekuertzt werden. Z.B. 4.787347812347e-124"""
evalue_first, evalue_last = str(hsp.expect).split('e')
evalue = str(round(float(evalue_first), 1)) + 'e' + evalue_last
else:
evalue = round(hsp.expect, 1)
"""
hit_def = change_according_reviewer(alignment.hit_def, note_line = True)
if hit_def.split('|')[:-1] != []:
hit_def = hit_def.split('|')[-1].split()[0]
else:
hit_def = accession
"""
hit_def = accession
"""
try:
protein_accession_gb = hit_def.split('gb|')[1].split('|')[0] #try to extract the genbank accession number >gi|302432474|gb|EFL04290.1|; -> EFL04290.1
inference = "similar to AA sequence:INSD: %s" % protein_accession_gb
feature_table_text[ hsp.bits ] += '\t\t\tinference\t%s\n' % (inference)
protein_accession_ref = hit_def.split('ref|')[1].split('|')[0] #try to extract the genbank accession number >gi|302432474|gb|EFL04290.1|; -> EFL04290.1
inference = "similar to AA sequence:RefSeq: %s" % protein_accession_ref
feature_table_text[ hsp.bits ] += '\t\t\tinference\t%s\n' % (inference)
except:
pass
"""
inference = """ab initio prediction:Augustus:2.5.5"""
feature_table_text[ hsp.bits ] += '\t\t\tinference\t%s\n' % (inference)
note = """similar to UniProtKB/Swiss-Prot Entry: %(hit_accession)s""" % {'gene_counter': gene_counter,
'accession':accession,
'alignment_hit_def': hit_def,
'hit_accession': alignment.accession,
'len': query_length,
'evalue': evalue,
'bit_score': round(hsp.bits, 2),
'locus_tag': locus_tag,
}
feature_table_text[ hsp.bits ] += '\t\t\tnote\t%s\n' % (note)
#for region in cds[1:]:
# mRNA_annotation += '%i\t%i\n' % (region.start, region.stop)
# feature_table.write('%i\t%i\n' % (region.start, region.stop))
if hsp_has_annotation == False:
"""
If hsp has no annotation, insert a hypothetical protein
"""
feature_table.write('%i\t%i\tgene\n' % (gene_start, gene_end))
feature_table.write('\t\t\tlocus_tag\t%s%04d\n' % (locus_tag, gene_counter))
assert cds[0].seq_type == 'CDS'
short_intron = check_short_introns(cds)
if short_intron:
feature_table.write('\t\t\tpseudo\n')
feature_table.write('\t\t\tnote\tnonfunctional\n')
"""
Write the CDS section for the 'no-annotation' case.
"""
feature_table.write('%i\t%i\tCDS\n' % (cds[0].start, cds[0].stop))
for region in cds[1:]:
feature_table.write('%i\t%i\n' % (region.start, region.stop))
feature_table.write('\t\t\tproduct\thypothetical protein\n')
feature_table.write('\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' % (locus_tag, gene_counter))
feature_table.write('\t\t\ttranscript_id\tgnl|PBUF|%smrna%04d\n' % (locus_tag, gene_counter))
feature_table.write('\t\t\tnote\tpredicted with Augustus 2.5.5\n')
"""
Write the mRNA section for the 'no-annotation' case.
"""
feature_table.write('%i\t%i\tmRNA\n' % (cds[0].start, cds[0].stop))
for region in cds[1:]:
feature_table.write('%i\t%i\n' % (region.start, region.stop))
feature_table.write('\t\t\tproduct\thypothetical protein\n')
feature_table.write('\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' % (locus_tag, gene_counter))
feature_table.write('\t\t\ttranscript_id\tgnl|PBUF|%smrna%04d\n' % (locus_tag, gene_counter))
else:
bitscores = feature_table_text.keys()
bitscores.sort(reverse=True)
feature_table.write(feature_table_text[ bitscores[0] ])
if feature_table_text[ bitscores[0] ].find('\t\t\tproduct\thypothetical protein\n') == -1:
annotation_count_with_putative_function += 1
return (gene_counter, annotation_count_with_putative_function)
def parse_blastxml(input_path, glimmer_mapping, feature_table, annotation_count_with_putative_function, gene_counter, locus_tag, min_coverage, min_ident):
# extract the sequence number
seq_number = input_path.split('Seq')[-1].split('.blastxml')[0]
feature_table.write('>Feature Seq%s\n' % seq_number)
with open(input_path) as blast_handle:
for entry in NCBIXML.parse(blast_handle):
if entry.application == "BLASTX":
query_length = entry.query_length
if type(query_length) == type(1.7):
print "Query length is not a multiple of three"
break
query_id = entry.query.split()[0]
query_info = glimmer_mapping[ query_id ]
query_start = int(query_info[0])
query_end = int(query_info[1])
else:
break
gene_counter += 1
"""
if not entry.alignments:
feature_table.write('%i\t%i\tgene\n' % (query_start, query_end))
feature_table.write('\t\t\tlocus_tag\t%s%04d\n' % (locus_tag, gene_counter))
feature_table.write('%i\t%i\tCDS\n' % (query_start, query_end))
feature_table.write('\t\t\tproduct\thypothetical protein\n')
feature_table.write('\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' % (locus_tag, gene_counter))
feature_table.write('\t\t\tnote\tpredicted with glimmer3\n')
break
"""
hsp_has_annotation = False
feature_table_text = dict()
for alignment in entry.alignments:
for hsp in alignment.hsps:
nident = hsp.identities
ident = (100*float(nident)/float(hsp.align_length))
coverage = False
if int(hsp.query_end) - int(hsp.query_start) + 1 >= min_coverage * query_length:
coverage = True
# only annotate hits with an identity over 50% and a coverage over 50%
if ident > min_ident and coverage:
feature_table_text[ hsp.bits ] = ""
hsp_has_annotation = True
accession = alignment.hit_def.split('OS=')[0].strip()
assert change_according_reviewer('Pimelyl-[acyl-carrier protein] methyl ester esterase', note_line = False) == 'Pimelyl-[acyl-carrier protein] methyl ester esterase'
assert change_according_reviewer('putative D-malate dehydrogenase [decarboxylating] [gnl|PBUF|STVIR_0046:1-352] [gnl|PBUF|STVIR_0046: raw, aa len= 352]', note_line = False) == 'D-malate dehydrogenase'
accession = change_according_reviewer(accession, note_line = False)
feature_table_text[ hsp.bits ] += '%i\t%i\tgene\n' % (query_start, query_end)
feature_table_text[ hsp.bits ] += '\t\t\tlocus_tag\t%s%04d\n' % (locus_tag, gene_counter)
feature_table_text[ hsp.bits ] += '%i\t%i\tCDS\n' % (query_start, query_end)
if accession.startswith('hypothetical protein') or \
accession.startswith('predicted protein') or \
accession == '' or accession == 'protein':
feature_table_text[ hsp.bits ] += '\t\t\tproduct\thypothetical protein\n'
else:
feature_table_text[ hsp.bits ] += '\t\t\tproduct\tputative %s\n' % (accession)
feature_table_text[ hsp.bits ] += '\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' % (locus_tag, gene_counter)
if str(hsp.expect).find('e') != -1:
""" Der evalue ist eine lange Zahl und muss gekuertzt werden. Z.B. 4.787347812347e-124"""
evalue_first, evalue_last = str(hsp.expect).split('e')
evalue = str(round(float(evalue_first), 1)) + 'e' + evalue_last
else:
evalue = round(hsp.expect, 1)
"""
hit_def = change_according_reviewer(alignment.hit_def, note_line = True)
if hit_def.split('|')[:-1] != []:
hit_def = hit_def.split('|')[-1].split()[0]
else:
hit_def = accession
"""
""""
try:
protein_accession_gb = hit_def.split('gb|')[1].split('|')[0] #try to extract the genbank accession number >gi|302432474|gb|EFL04290.1|; -> EFL04290.1
inference = "similar to AA sequence:INSD: %s" % protein_accession_gb
feature_table_text[ hsp.bits ] += '\t\t\tinference\t%s\n' % (inference)
protein_accession_ref = hit_def.split('ref|')[1].split('|')[0] #try to extract the genbank accession number >gi|302432474|gb|EFL04290.1|; -> EFL04290.1
inference = "similar to AA sequence:RefSeq: %s" % protein_accession_ref
feature_table_text[ hsp.bits ] += '\t\t\tinference\t%s\n' % (inference)
except:
pass
"""
inference = """ab initio prediction:Glimmer:3"""
feature_table_text[ hsp.bits ] += '\t\t\tinference\t%s\n' % (inference)
note = """similar to UniProtKB/Swiss-Prot Entry: %(hit_accession)s""" % {'gene_counter': gene_counter,
'accession':accession,
'alignment_hit_def': accession,
'hit_accession': alignment.accession,
'len': query_length,
'evalue': evalue,
'bit_score': round(hsp.bits, 2),
'locus_tag': locus_tag,
}
feature_table_text[ hsp.bits ] += '\t\t\tnote\t%s\n' % (note)
if hsp_has_annotation == False:
"""
If hsp has no annotation with the specified identity and coverage, insert a hypothetical protein
"""
feature_table.write('%i\t%i\tgene\n' % (query_start, query_end))
feature_table.write('\t\t\tlocus_tag\t%s%04d\n' % (locus_tag, gene_counter))
feature_table.write('%i\t%i\tCDS\n' % (query_start, query_end))
feature_table.write('\t\t\tproduct\thypothetical protein\n')
feature_table.write('\t\t\tprotein_id\tgnl|PBUF|%s%04d\n' % (locus_tag, gene_counter))
feature_table.write('\t\t\tnote\tab initio prediction:Glimmer3\n')
else:
bitscores = feature_table_text.keys()
bitscores.sort(reverse=True)
feature_table.write(feature_table_text[ bitscores[0] ])
if feature_table_text[ bitscores[0] ].find('\t\t\tproduct\thypothetical protein\n') == -1:
annotation_count_with_putative_function += 1
return (gene_counter, annotation_count_with_putative_function)