def add_aragorn_features(assemblies, features, aragorn_file):
current_assembly_id = None
for line in open(aragorn_file):
line = line.rstrip()
if line.startswith('>'):
m = re.match('>(\S+)', line)
current_assembly_id = m.group(1)
if current_assembly_id not in assemblies:
assembly = things.Assembly(id=current_assembly_id, residues='')
assemblies[current_assembly_id] = assembly
else:
cols = line.split()
if len(cols) == 5:
if cols[1].startswith('tRNA'):
feat_type = 'tRNA'
elif cols[1].startswith('tmRNA'):
feat_type = 'tmRNA'
elif cols[1].startswith('mtRNA'):
feat_type = 'mtRNA'
else:
raise Exception("Unexpected type in ARAGORN, column value: {0}".format(cols[1]))
feat_base = "{0}_{1}".format(feat_type, uuid.uuid4())
gene_id = "{0}_gene".format(feat_base)
RNA_id = "{0}_{1}".format(feat_base, feat_type)
m = re.match('(c*)\[(\d+),(\d+)\]', cols[2])
if m:
rfmin = int(m.group(2)) - 1
rfmax = int(m.group(3))
if m.group(1):
rstrand = -1
else:
rstrand = 1
else:
raise Exception("ERROR: unexpected coordinate format: {0}".format(cols[2]))
current_assembly = assemblies[current_assembly_id]
gene = things.Gene(id=gene_id)
gene.locate_on(target=current_assembly, fmin=rfmin, fmax=rfmax, strand=rstrand)
features[gene_id] = gene
current_assembly.add_gene(gene)
if feat_type == 'tRNA':
RNA = things.tRNA(id=RNA_id, parent=gene, anticodon=cols[4][1:4].upper())
gene.add_tRNA(RNA)
else:
RNA = things.tmRNA(id=RNA_id, parent=gene)
gene.add_tmRNA(RNA)
RNA.locate_on(target=current_assembly, fmin=rfmin, fmax=rfmax, strand=rstrand)
RNA.annotation = annotation.FunctionalAnnotation(product_name=cols[1])
features[RNA_id] = RNA
def process_gb_gene(gene):
annot = annotation.FunctionalAnnotation()
annot.other_attributes['ensembl_version'] = gene.qualifiers['gene'][0]
m = re.match('(.+)\.\d+', annot.other_attributes['ensembl_version'])
if m:
annot.other_attributes['ensembl_id'] = m.group(1)
else:
raise Exception("Found an ensembl id ({0}) without a .version".format(
annot.other_attributes['ensembl_version']))
annot.gene_symbol = get_gene_sym(gene)
annot.product_name = get_product(gene)
annot.other_attributes['loc'] = get_coordinates(gene)
return annot
def initialize_polypeptides( log_fh, fasta_file ):
'''
Reads a FASTA file of (presumably) polypeptide sequences and creates a dict of Polypeptide
objects, keyed by ID, with bioannotation.FunctionalAnnotation objects attached.
'''
seqs = utils.fasta_dict_from_file(fasta_file)
polypeptides = dict()
for seq_id in seqs:
polypeptide = things.Polypeptide(id=seq_id, length=len(seqs[seq_id]['s']), residues=seqs[seq_id]['s'])
annot = annotation.FunctionalAnnotation(product_name=DEFAULT_PRODUCT_NAME)
log_fh.write("INFO: {0}: Set initial product name to '{1}'\n".format(seq_id, DEFAULT_PRODUCT_NAME))
polypeptide.annotation = annot
polypeptides[seq_id] = polypeptide
return polypeptides
def main():
parser = argparse.ArgumentParser(
description='Convert GenBank flat files to GFF3 format')
## output file to be written
parser.add_argument('-i',
'--input_file',
type=str,
required=True,
help='Path to an input GBK file')
parser.add_argument('-o',
'--output_file',
type=str,
required=False,
help='Path to an output GFF file to be created')
parser.add_argument(
'--with_fasta',
dest='fasta',
action='store_true',
help=
'Include the FASTA section with genomic sequence at end of file. (default)'
)
parser.add_argument('--no_fasta', dest='fasta', action='store_false')
parser.set_defaults(fasta=True)
args = parser.parse_args()
## output will either be a file or STDOUT
ofh = sys.stdout
if args.output_file is not None:
ofh = open(args.output_file, 'wt')
ofh.write("##gff-version 3\n")
assemblies = dict()
current_assembly = None
current_gene = None
current_RNA = None
rna_count_by_gene = defaultdict(int)
exon_count_by_RNA = defaultdict(int)
seqs_pending_writes = False
features_skipped_count = 0
# each gb_record is a SeqRecord object
for gb_record in SeqIO.parse(open(args.input_file, "r"), "genbank"):
mol_id = gb_record.name
if mol_id not in assemblies:
assemblies[mol_id] = things.Assembly(id=mol_id)
if len(str(gb_record.seq)) > 0:
seqs_pending_writes = True
assemblies[mol_id].residues = str(gb_record.seq)
assemblies[mol_id].length = len(str(gb_record.seq))
current_assembly = assemblies[mol_id]
# each feat is a SeqFeature object
for feat in gb_record.features:
#print(feat)
fmin = int(feat.location.start)
fmax = int(feat.location.end)
if feat.location.strand == 1:
strand = '+'
elif feat.location.strand == -1:
strand = '-'
else:
raise Exception(
"ERROR: unstranded feature encountered: {0}".format(feat))
#print("{0} located at {1}-{2} strand:{3}".format( locus_tag, fmin, fmax, strand ) )
if feat.type == 'source':
continue
if feat.type == 'gene':
# print the previous gene (if there is one)
if current_gene is not None:
gene.print_as(fh=ofh, source='GenBank', format='gff3')
locus_tag = feat.qualifiers['locus_tag'][0]
gene = things.Gene(id=locus_tag, locus_tag=locus_tag)
gene.locate_on(target=current_assembly,
fmin=fmin,
fmax=fmax,
strand=strand)
current_gene = gene
current_RNA = None
elif feat.type == 'mRNA':
locus_tag = feat.qualifiers['locus_tag'][0]
rna_count_by_gene[locus_tag] += 1
feat_id = "{0}.mRNA.{1}".format(locus_tag,
rna_count_by_gene[locus_tag])
mRNA = things.mRNA(id=feat_id,
parent=current_gene,
locus_tag=locus_tag)
mRNA.locate_on(target=current_assembly,
fmin=fmin,
fmax=fmax,
strand=strand)
gene.add_mRNA(mRNA)
current_RNA = mRNA
if feat_id in exon_count_by_RNA:
raise Exception(
"ERROR: two different RNAs found with same ID: {0}".
format(feat_id))
else:
exon_count_by_RNA[feat_id] = 0
elif feat.type == 'tRNA':
locus_tag = feat.qualifiers['locus_tag'][0]
rna_count_by_gene[locus_tag] += 1
feat_id = "{0}.tRNA.{1}".format(locus_tag,
rna_count_by_gene[locus_tag])
if 'product' in feat.qualifiers:
anticodon = feat.qualifiers['product'][0]
else:
anticodon = None
tRNA = things.tRNA(id=feat_id,
parent=current_gene,
anticodon=anticodon)
tRNA.locate_on(target=current_assembly,
fmin=fmin,
fmax=fmax,
strand=strand)
gene.add_tRNA(tRNA)
current_RNA = tRNA
if feat_id in exon_count_by_RNA:
raise Exception(
"ERROR: two different RNAs found with same ID: {0}".
format(feat_id))
else:
exon_count_by_RNA[feat_id] = 0
elif feat.type == 'rRNA':
locus_tag = feat.qualifiers['locus_tag'][0]
rna_count_by_gene[locus_tag] += 1
feat_id = "{0}.rRNA.{1}".format(locus_tag,
rna_count_by_gene[locus_tag])
if 'product' in feat.qualifiers:
product = feat.qualifiers['product'][0]
else:
product = None
annot = annotation.FunctionalAnnotation(product_name=product)
rRNA = things.rRNA(id=feat_id,
parent=current_gene,
annotation=annot)
rRNA.locate_on(target=current_assembly,
fmin=fmin,
fmax=fmax,
strand=strand)
gene.add_rRNA(rRNA)
current_RNA = rRNA
if feat_id in exon_count_by_RNA:
raise Exception(
"ERROR: two different RNAs found with same ID: {0}".
format(feat_id))
else:
exon_count_by_RNA[feat_id] = 0
elif feat.type == 'CDS':
locus_tag = feat.qualifiers['locus_tag'][0]
# If processing a prokaryotic GBK, we'll encounter CDS before mRNA, so we have to
# manually make one
if current_RNA is None:
feat_id = "{0}.mRNA.{1}".format(
locus_tag, rna_count_by_gene[locus_tag])
mRNA = things.mRNA(id=feat_id, parent=current_gene)
mRNA.locate_on(target=current_assembly,
fmin=fmin,
fmax=fmax,
strand=strand)
gene.add_mRNA(mRNA)
current_RNA = mRNA
if 'product' in feat.qualifiers:
product = feat.qualifiers['product'][0]
else:
product = None
if 'gene' in feat.qualifiers:
gene_symbol = feat.qualifiers['gene'][0]
else:
gene_symbol = None
annot = annotation.FunctionalAnnotation(
product_name=product, gene_symbol=gene_symbol)
if 'db_xref' in feat.qualifiers:
for dbxref in feat.qualifiers['db_xref']:
annot.add_dbxref(dbxref)
polypeptide_id = "{0}.polypeptide.{1}".format(
locus_tag, rna_count_by_gene[locus_tag])
polypeptide = things.Polypeptide(id=polypeptide_id,
parent=mRNA,
annotation=annot)
mRNA.add_polypeptide(polypeptide)
exon_count_by_RNA[current_RNA.id] += 1
cds_id = "{0}.CDS.{1}".format(
current_RNA.id, exon_count_by_RNA[current_RNA.id])
current_CDS_phase = 0
for loc in feat.location.parts:
subfmin = int(loc.start)
subfmax = int(loc.end)
CDS = things.CDS(id=cds_id, parent=current_RNA)
CDS.locate_on(target=current_assembly,
fmin=subfmin,
fmax=subfmax,
strand=strand,
phase=current_CDS_phase)
current_RNA.add_CDS(CDS)
# calculate the starting phase for the next CDS feature (in case there is one)
# 0 + 6 = 0 TTGCAT
# 0 + 7 = 2 TTGCATG
# 1 + 6 = 1 TTGCAT
# 2 + 7 = 1 TTGCATG
# general: 3 - ((length - previous phase) % 3)
current_CDS_phase = 3 - ((
(subfmax - subfmin) - current_CDS_phase) % 3)
if current_CDS_phase == 3:
current_CDS_phase = 0
exon_id = "{0}.exon.{1}".format(
current_RNA.id, exon_count_by_RNA[current_RNA.id])
exon = things.Exon(id=exon_id, parent=current_RNA)
exon.locate_on(target=current_assembly,
fmin=subfmin,
fmax=subfmax,
strand=strand)
current_RNA.add_exon(exon)
exon_count_by_RNA[current_RNA.id] += 1
else:
print(
"WARNING: The following feature was skipped:\n{0}".format(
feat))
features_skipped_count += 1
# don't forget to do the last gene, if there were any
if current_gene is not None:
gene.print_as(fh=ofh, source='GenBank', format='gff3')
if args.fasta is True:
if seqs_pending_writes is True:
ofh.write("##FASTA\n")
for assembly_id in assemblies:
ofh.write(">{0}\n".format(assembly_id))
ofh.write("{0}\n".format(
utils.wrapped_fasta(assemblies[assembly_id].residues)))
if features_skipped_count > 0:
print("Warning: {0} unsupported feature types were skipped".format(
features_skipped_count))
def add_aragorn_features(assemblies, features, aragorn_file):
current_assembly_id = None
for line in open(aragorn_file):
line = line.rstrip()
if line.startswith('>'):
m = re.match('>(\S+)', line)
current_assembly_id = m.group(1)
if current_assembly_id not in assemblies:
assembly = things.Assembly(id=current_assembly_id, residues='')
assemblies[current_assembly_id] = assembly
else:
cols = line.split()
if len(cols) == 5:
if cols[1].startswith('tRNA'):
feat_type = 'tRNA'
elif cols[1].startswith('tmRNA'):
feat_type = 'tmRNA'
elif cols[1].startswith('mtRNA'):
feat_type = 'mtRNA'
else:
raise Exception("Unexpected type in ARAGORN, column value: {0}".format(cols[1]))
feat_base = "{0}_{1}".format(feat_type, uuid.uuid4())
gene_id = "{0}_gene".format(feat_base)
RNA_id = "{0}_{1}".format(feat_base, feat_type)
m = re.match('(c*)\[(\d+),(\d+)\]', cols[2])
if m:
rfmin = int(m.group(2)) - 1
rfmax = int(m.group(3))
# For predictions spanning the origin of circular molecules, fmax needs to be adjusted
# https://github.com/The-Sequence-Ontology/Specifications/blob/master/gff3.md
# Sub-heading: Circular Genomes
#
# Example input lines:
# 62 tRNA-Met c[2764659,18] 29 (cat)
# 1 tRNA-Ile [2697442,47] 35 (gat)
if rfmax < rfmin:
assemblies[current_assembly_id].is_circular = True
rfmax = rfmin + rfmax + 1
if m.group(1):
rstrand = -1
else:
rstrand = 1
else:
raise Exception("ERROR: unexpected coordinate format: {0}".format(cols[2]))
current_assembly = assemblies[current_assembly_id]
gene = things.Gene(id=gene_id)
gene.locate_on(target=current_assembly, fmin=rfmin, fmax=rfmax, strand=rstrand)
features[gene_id] = gene
current_assembly.add_gene(gene)
if feat_type == 'tRNA':
RNA = things.tRNA(id=RNA_id, parent=gene, anticodon=cols[4][1:4].upper())
gene.add_tRNA(RNA)
else:
RNA = things.tmRNA(id=RNA_id, parent=gene)
gene.add_tmRNA(RNA)
RNA.locate_on(target=current_assembly, fmin=rfmin, fmax=rfmax, strand=rstrand)
RNA.annotation = annotation.FunctionalAnnotation(product_name=cols[1])
features[RNA_id] = RNA