def _record_with_genes_only(reference):
gene_names = Transcript.query.with_entities(Transcript.gene). \
filter_by(reference_id=reference.id).all()
record = _bare_record(reference)
genes = []
for gene_name in gene_names:
gene = Gene(gene_name.gene)
version = gene.newLocusTag()
my_transcript = Locus(version)
my_transcript.mRNA = PList()
my_transcript.CDS = PList()
gene.transcriptList.append(my_transcript)
genes.append(gene)
record.geneList = genes
record.seq = Seq('', generic_dna)
return record
def _get_mutalyzer_record(reference, db_transcripts):
"""
Creates a Mutalyzer specific record from the transcript entries retrieved
from the gbparser database.
:param reference: A gbparser database reference entry.
:param db_transcripts:A gbparser database list of transcript.
:return: The Mutalyzer record.
"""
record = _bare_record(reference)
# Extracting the transcripts from the DB entries.
transcripts = []
for db_transcript in db_transcripts:
transcript = {
'gene':
db_transcript.gene,
'strand':
db_transcript.strand,
'transcript_start':
db_transcript.transcript_start,
'transcript_stop':
db_transcript.transcript_stop,
'transcript_product':
db_transcript.transcript_product,
'exons': [],
'exons_start':
db_transcript.exons_start,
'exons_stop':
db_transcript.exons_stop,
'transcriptID':
db_transcript.transcript_accession + '.' +
db_transcript.transcript_version,
}
if db_transcript.protein_accession is not None \
and db_transcript.protein_version is not None:
transcript['cds_start'] = db_transcript.cds_start
transcript['cds_stop'] = db_transcript.cds_stop
transcript['protein_product'] = db_transcript.protein_product
transcript['proteinID'] = '%s.%s' %\
(db_transcript.protein_accession,
db_transcript.protein_version)
transcript['linkMethod'] = 'ncbi'
starts = map(int, db_transcript.exons_start.split(',')) \
if db_transcript.exons_start else None
stops = map(int, db_transcript.exons_stop.split(',')) \
if db_transcript.exons_stop else None
if (starts and stops) and (len(starts) == len(stops)):
for start, stop in zip(starts, stops):
exon = {'start': start, 'stop': stop}
transcript['exons'].append(exon)
transcripts.append(transcript)
# Generating the actual record entries in the Mutalyzer format.
gene_dict = {}
for db_transcript in transcripts:
if db_transcript['gene'] in gene_dict:
gene = gene_dict[db_transcript['gene']]
else:
gene = Gene(db_transcript['gene'])
if db_transcript['strand'] == '+':
gene.orientation = 1
if db_transcript['strand'] == '-':
gene.orientation = -1
transcript = Locus(gene.newLocusTag())
transcript.mRNA = PList()
transcript.mRNA.location = [
db_transcript['transcript_start'], db_transcript['transcript_stop']
]
transcript.transcriptID = db_transcript['transcriptID']
transcript.exon = PList()
if db_transcript.get('exons') \
and isinstance(db_transcript.get('exons'), list):
exon_list = []
for exon in db_transcript['exons']:
exon_list.extend([exon['start'], exon['stop']])
transcript.exon.positionList = exon_list
else:
transcript.exon.positionList = transcript.mRNA.location
transcript.mRNA.positionList = transcript.exon.positionList
transcript.mRNA.positionList.sort()
if db_transcript.get('proteinID'):
transcript.CDS = PList()
transcript.CDS.location = [
db_transcript['cds_start'], db_transcript['cds_stop']
]
transcript.CDS.positionList = cds_position_list(
transcript.mRNA.positionList, transcript.CDS.location)
transcript.proteinID = db_transcript['proteinID']
transcript.transcriptProduct = db_transcript['transcript_product']
transcript.proteinProduct = db_transcript['protein_product']
transcript.linkMethod = 'ncbi'
transcript.transcribe = True
transcript.translate = True
else:
transcript.linkMethod = None
transcript.transcribe = True
transcript.translate = False
transcript.locusTag = ''
# transcript.molType = db_transcript['molType']
gene.transcriptList.append(transcript)
gene_dict[gene.name] = gene
record.geneList = list(gene_dict.values())
# Get the sequence.
seq_path = settings.SEQ_PATH + reference.checksum_sequence + '.sequence'
try:
seq = Seq(_get_sequence_mmap(seq_path, 1, reference.length + 1),
generic_dna)
except IOError:
return None
else:
record.seq = seq
return record
def create_record(self, filename):
"""
Create a GenRecord.Record from a GenBank file
@arg filename: The full path to the compressed GenBank file
@type filename: unicode
@return: A GenRecord.Record instance
@rtype: object (record)
"""
# first create an intermediate genbank record with BioPython
file_handle = bz2.BZ2File(filename, "r")
file_handle = codecs.getreader('utf-8')(file_handle)
biorecord = SeqIO.read(file_handle, "genbank")
file_handle.close()
record = Record()
record.seq = biorecord.seq
# Note: The .source_* values may be different from the values we are
# working with, e.g. for UD slices where these values (taken from
# the genbank file) are from the original NC reference. We try to
# set the .id field to the working value in the caller.
record.source_id = biorecord.id
record.source_accession, record.source_version = biorecord.id.split(
'.')[:2]
record.source_gi = biorecord.annotations['gi']
record.organism = biorecord.annotations['organism']
# Todo: This will change once we support protein references
if isinstance(biorecord.seq.alphabet, ProteinAlphabet):
return record
exonList = []
geneDict = {}
accInfo = biorecord.annotations['accessions']
if len(accInfo) >= 3 and accInfo[1] == "REGION:":
# Todo: This information is present in the genbank file if it is a
# UD sliced from a chromosome. We can get the same information
# for NM references from our mapping database and that way
# also provide chromosomal variant descriptions for those.
region = accInfo[2]
if "complement" in region:
record.orientation = -1
record.chromOffset = int(region.split('.')[2][:-1])
#if
else:
record.chromOffset = int(accInfo[2].split('.')[0])
#if
for i in biorecord.features:
if i.qualifiers:
if i.type == "source":
if i.qualifiers.has_key("mol_type"):
if i.qualifiers["mol_type"][0] in ["mRNA", \
"transcribed RNA"] :
record.molType = 'n'
else:
record.molType = 'g'
#if
if i.qualifiers.has_key("organelle"):
record.organelle = i.qualifiers["organelle"][0]
if record.organelle == "mitochondrion":
record.molType = 'm'
#if
fakeGene = Locus("001")
record.source.transcriptList.append(fakeGene)
fakeGene.CDS = PList()
fakeGene.CDS.location = self.__location2pos(i.location)
#if
if i.qualifiers.has_key("gene"):
if not unicode(i.location.start).isdigit() or \
not unicode(i.location.end).isdigit():
# Feature is not completely in reference. Either start
# or end is not a Bio.SeqFeature.ExactPosition.
continue
geneName = i.qualifiers["gene"][0]
if i.type == "gene":
if not geneDict.has_key(geneName):
myGene = Gene(geneName)
record.geneList.append(myGene)
if i.strand:
myGene.orientation = i.strand
myGene.location = self.__location2pos(i.location)
geneDict[geneName] = tempGene(geneName)
#if
else:
if geneName not in geneDict:
# We should have seen a gene entry for this gene
# by now. Could be that it was skipped because it
# was not completely in reference (see check
# above). In that case we just ignore any of its
# features.
continue
#if
if i.type in [
"mRNA", "misc_RNA", "ncRNA", "rRNA", "tRNA",
"tmRNA"
]:
geneDict[geneName].rnaList.append(i)
if i.type == "CDS":
geneDict[geneName].cdsList.append(i)
if i.type == "exon":
exonLocation = self.__location2pos(i.location)
if exonLocation:
exonList.extend(exonLocation)
#if
#if
#if
#for
if record.molType in ['g', 'm']:
for j in geneDict.keys():
myGene = geneDict[j]
self.link(myGene.rnaList, myGene.cdsList)
for i in myGene.rnaList:
if i.usable:
myRealGene = record.findGene(i.gene)
if i.locus_tag:
# Note: We use the last three characters of the
# locus_tag as a unique transcript version id.
# This is also used to for the protein-transcript
# link table.
# Normally, locus_tag ends with three digits, but
# for some (e.g. mobA on NC_011228, a plasmid) it
# ends with two digits prepended with an
# underscore. Or prepended with a letter. We
# really want a number, so 'fix' this by only
# looking for a numeric part.
try:
version = LOCUS_TAG_VERSION.findall(
i.locus_tag)[0].zfill(3)
except IndexError:
version = '000'
myTranscript = Locus(version)
else:
myTranscript = Locus(myRealGene.newLocusTag())
myTranscript.mRNA = PList()
myTranscript.mRNA.positionList = i.positionList
myTranscript.mRNA.location = i.location
myTranscript.transcribe = True
myTranscript.transcriptID = i.transcript_id
myTranscript.transcriptProduct = i.product
myTranscript.locusTag = i.locus_tag
if i.link:
myTranscript.CDS = PList()
myTranscript.CDS.positionList = i.link.positionList
myTranscript.CDS.location = i.link.location
myTranscript.translate = True
myTranscript.proteinID = i.link.protein_id
myTranscript.linkMethod = i.linkMethod
myTranscript.proteinProduct = i.link.product
if i.link.qualifiers.has_key("transl_table"):
myTranscript.txTable = \
int(i.qualifiers["transl_table"][0])
#if
myRealGene.transcriptList.append(myTranscript)
#if
#for
for i in myGene.cdsList:
if not i.linked and \
(i.usable or not geneDict[myGene.name].rnaList) :
myRealGene = record.findGene(i.gene)
if i.locus_tag:
# Note: We use the last three characters of the
# locus_tag as a unique transcript version id.
# This is also used to for the protein-transcript
# link table.
# Normally, locus_tag ends with three digits, but
# for some (e.g. mobA on NC_011228, a plasmid) it
# ends with two digits prepended with an
# underscore. Or prepended with a letter. We
# really want a number, so 'fix' this by only
# looking for a numeric part.
try:
version = LOCUS_TAG_VERSION.findall(
i.locus_tag)[0].zfill(3)
except IndexError:
version = '000'
myTranscript = Locus(version)
else:
myTranscript = Locus(myRealGene.newLocusTag())
myTranscript.CDS = PList()
myTranscript.CDS.positionList = i.positionList
myTranscript.CDS.location = i.location
myTranscript.proteinID = i.protein_id
myTranscript.proteinProduct = i.product
if i.qualifiers.has_key("transl_table"):
myTranscript.txTable = \
int(i.qualifiers["transl_table"][0])
myRealGene.transcriptList.append(myTranscript)
#if
#if
#for
#for
#if
else:
if geneDict:
myGene = geneDict[geneDict.keys()[0]]
myRealGene = record.geneList[0]
if myGene.cdsList:
myCDS = myGene.cdsList[0]
self.__tagByDict(myCDS, "protein_id")
self.__tagByDict(myCDS, "product")
#if
else:
myCDS = None
myTranscript = Locus("001")
myTranscript.exon = PList()
if exonList:
myTranscript.exon.positionList = exonList
else:
myTranscript.exon.location = myRealGene.location
if myCDS:
myTranscript.CDS = PList()
myTranscript.CDS.location = \
self.__location2pos(myCDS.location)
#if
if exonList or myRealGene.location or \
myTranscript.CDS.location :
myTranscript.transcriptID = biorecord.id
if myCDS:
myTranscript.proteinID = myCDS.protein_id
myTranscript.proteinProduct = myCDS.product
myTranscript.linkMethod = "exhaustion"
myTranscript.transcribe = True
if myCDS.qualifiers.has_key("transl_table"):
myTranscript.txTable = \
int(i.qualifiers["transl_table"][0])
#if
myRealGene.transcriptList.append(myTranscript)
#if
#if
#else
for i in record.geneList:
if not i.transcriptList:
record.geneList.remove(i)
return record
def _get_mutalyzer_record(reference, db_transcripts):
"""
Creates a Mutalyzer specific record from the transcript entries retrieved
from the gbparser database.
:param reference: A gbparser database reference entry.
:param db_transcripts:A gbparser database list of transcript.
:return: The Mutalyzer record.
"""
record = _bare_record(reference)
# Extracting the transcripts from the DB entries.
transcripts = []
for transcript in db_transcripts:
my_transcript = {
'gene': transcript.gene,
'strand': transcript.strand,
'transcript_start': transcript.transcript_start,
'transcript_stop': transcript.transcript_stop,
'cds_start': transcript.cds_start,
'cds_stop': transcript.cds_stop,
'exons': [],
'exons_start': transcript.exons_start,
'exons_stop': transcript.exons_stop,
'transcriptID': transcript.transcript_accession + '.' +
transcript.transcript_version,
'proteinID':
transcript.protein_accession + '.' + transcript.protein_version,
'linkMethod': 'ncbi'
}
# if transcript.exons_start:
# starts = transcript.exons_start.split(',')
# if transcripts.exons_stop:
# stops = transcript.exons_stopts.split(',')
starts = map(int, transcript.exons_start.split(
',')) if transcript.exons_start else None
stops = map(int, transcript.exons_stop.split(
',')) if transcript.exons_stop else None
if (starts and stops) and (len(starts) == len(stops)):
for start, stop in zip(starts, stops):
exon = {'start': start, 'stop': stop}
my_transcript['exons'].append(exon)
# if transcript.exons and isinstance(transcript.exons, list):
# for exon in transcript.exons:
# exon = {'start': exon.start,
# 'stop': exon.stop}
# my_transcript['exons'].append(exon)
transcripts.append(my_transcript)
# Generating the actual record entries in the Mutalyzer format.
gene_dict = {}
for transcript in transcripts:
if transcript['gene'] in gene_dict:
gene = gene_dict[transcript['gene']]
else:
gene = Gene(transcript['gene'])
if transcript['strand'] == '+':
gene.orientation = 1
if transcript['strand'] == '-':
gene.orientation = -1
my_transcript = Locus(gene.newLocusTag())
my_transcript.mRNA = PList()
my_transcript.mRNA.location = [
transcript['transcript_start'], transcript['transcript_stop']
]
my_transcript.CDS = PList()
my_transcript.CDS.location = [
transcript['cds_start'], transcript['cds_stop']
]
my_transcript.exon = PList()
if transcript.get('exons') and isinstance(transcript.get('exons'),
list):
exon_list = []
for exon in transcript['exons']:
exon_list.extend([exon['start'], exon['stop']])
my_transcript.exon.positionList = exon_list
else:
my_transcript.exon.positionList = my_transcript.mRNA.location
my_transcript.mRNA.positionList = my_transcript.exon.positionList
my_transcript.mRNA.positionList.sort()
my_transcript.CDS.positionList = cds_position_list(
my_transcript.mRNA.positionList, my_transcript.CDS.location)
my_transcript.transcriptID = transcript['transcriptID']
my_transcript.proteinID = transcript['proteinID']
my_transcript.linkMethod = 'ncbi'
my_transcript.transcribe = True
my_transcript.translate = True
gene.transcriptList.append(my_transcript)
gene_dict[gene.name] = gene
record.geneList = list(gene_dict.values())
# Get the sequence.
seq_path = settings.SEQ_PATH + reference.checksum_sequence + '.sequence'
try:
seq = Seq(_get_sequence_mmap(seq_path, 1, reference.length + 1),
generic_dna)
except IOError:
return None
else:
record.seq = seq
return record