def translate_mvf(args):
"""Main method"""
args.qprint("Running TranslateMVF")
if args.gff:
args.qprint("Reading and Indexing MVF.")
else:
args.qprint("Reading MVF.")
mvf = MultiVariantFile(args.mvf, 'read', contigindex=bool(args.gff))
if mvf.flavor != 'dna':
raise RuntimeError("MVF must be flavor=dna to translate")
if args.gff:
args.qprint("Processing MVF Index File.")
mvf.read_index_file()
args.qprint("GFF processing start.")
gff_genes, gene_order = parse_gff_exome(args)
args.qprint("GFF processed.")
outmvf = MultiVariantFile(args.out, 'write', overwrite=args.overwrite)
outmvf.copy_headers_from(mvf)
outmvf.contig_data = dict(
(
i, dict((y, z)
for (y, z) in gff_genes[x].items()
if y not in ('cds', )))
for (i, x) in enumerate(gene_order))
outmvf.contig_indices = list(range(len(gene_order)))
outmvf.contig_ids = [gff_genes[x]['id']
for x in gene_order]
outmvf.contig_labels = [gff_genes[x]['label']
for x in gene_order]
outmvf.flavor = args.output_data
outmvf.metadata.notes.append(args.command_string)
outmvf.write_data(outmvf.get_header())
args.qprint("Output MVF Established.")
entrybuffer = []
nentry = 0
pos = None
if not args.gff:
args.qprint("No GFF used, translating sequences as pre-aligned in "
"coding frame.")
inputbuffer = []
current_contig = ''
for contigid, pos, allelesets in mvf.iterentries(decode=False):
if current_contig == '':
current_contig = contigid[:]
if contigid == current_contig:
inputbuffer.append((pos, allelesets))
else:
for _, amino_acids, alleles in iter_codons(
inputbuffer, mvf):
if all([x in '-X' for x in amino_acids]):
continue
if args.output_data == 'protein':
entrybuffer.append(
(current_contig, pos, (amino_acids,)))
else:
entrybuffer.append((
current_contig, pos, (
amino_acids, alleles[0],
alleles[1], alleles[2])))
nentry += 1
if nentry == args.line_buffer:
outmvf.write_entries(entrybuffer)
entrybuffer = []
nentry = 0
inputbuffer = [(pos, allelesets)]
current_contig = contigid[:]
if inputbuffer:
for _, amino_acids, alleles in iter_codons(
inputbuffer, outmvf):
if all([x in '-X' for x in amino_acids]):
continue
if args.output_data == 'protein':
entrybuffer.append(
(current_contig, pos, (amino_acids,)))
else:
entrybuffer.append((
current_contig, pos, (
amino_acids, alleles[0],
alleles[1], alleles[2])))
nentry += 1
if nentry == args.line_buffer:
outmvf.write_entries(entrybuffer)
entrybuffer = []
nentry = 0
else:
running_gene_index = -1
for igene, gene in enumerate(gene_order):
xcontiglabel = gff_genes[gene]['contig']
xcontig = mvf.get_contig_indices(
labels=gff_genes[gene]['contig'])
if xcontig is None:
print("Warning: contig {} not found".format(
gff_genes[gene]['contig']))
xcontigid = mvf.get_contig_ids(indices=xcontig)[0]
min_gene_coord = gff_genes[gene]['cds'][0][0]
max_gene_coord = gff_genes[gene]['cds'][-1][1]
mvf_entries = {}
if not igene % 100:
args.qprint("Processing gene {} on {}".format(
gene, xcontiglabel))
for contigid, pos, allelesets in mvf.itercontigentries(
xcontig, decode=False):
if pos < min_gene_coord:
continue
if pos > max_gene_coord:
break
mvf_entries[pos] = allelesets[0]
reverse_strand = gff_genes[gene]['strand'] == '-'
coords = []
running_gene_index += 1
for elem in gff_genes[gene]['cds']:
coords.extend(list(range(elem[0], elem[1] + 1)))
if reverse_strand:
coords = coords[::-1]
for codoncoord in range(0, len(coords), 3):
alleles = tuple(mvf_entries.get(x, '-')
for x in coords[codoncoord:codoncoord + 3])
if len(alleles) < 3:
alleles = tuple(list(alleles) + ['-'] * (3 - len(alleles)))
if all(len(x) == 1 for x in alleles):
if reverse_strand:
alleles = tuple(
MLIB.complement_bases[x] for x in alleles)
decoded_alleles = alleles
amino_acids = translate_single_codon(''.join(alleles))
else:
if reverse_strand is True:
decoded_alleles = tuple(tuple(MLIB.complement_bases[y]
for y in mvf.decode(x))
for x in alleles)
alleles = tuple(outmvf.encode(''.join(x))
for x in decoded_alleles)
else:
decoded_alleles = tuple(mvf.decode(x) for x in alleles)
amino_acids = tuple(translate_single_codon(''.join(x))
for x in zip(*decoded_alleles))
amino_acids = outmvf.encode(''.join(amino_acids))
if args.output_data == 'protein':
entrybuffer.append((
(
xcontigid
if args.retain_contigs
else running_gene_index
),
(
coords[codoncoord]
if args.retain_coords
else codoncoord
),
(
amino_acids,
)
))
elif args.output_data == 'codon':
entrybuffer.append((
(
xcontigid
if args.retain_contigs
else running_gene_index
),
(
coords[codoncoord]
if args.retain_coords
else codoncoord
),
(
amino_acids,
alleles[0],
alleles[1],
alleles[2]
)
))
elif args.output_data == 'dna':
for j, elem in enumerate(
range(codoncoord,
min(codoncoord + 3, len(coords)))):
entrybuffer.append((
(
xcontigid
if args.retain_contigs
else running_gene_index
),
(
coords[elem]
if args.retain_coords
else elem + 1
),
(
alleles[j],
)
))
nentry += 1
if nentry >= args.line_buffer:
args.qprint("Writing a block of {} entries.".format(
args.line_buffer))
outmvf.write_entries(entrybuffer)
entrybuffer = []
nentry = 0
if entrybuffer:
outmvf.write_entries(entrybuffer)
entrybuffer = []
nentry = 0
return ''
def merge_mvf(args):
"""Main method"""
args.qprint("Running MergeMVF")
if any(fpath.endswith('.gz') for fpath in args.mvf):
print("WARNING! Running MergeMVF with gzipped input files is "
"extremely slow and strongly discouraged.")
concatmvf = MultiVariantFile(args.out, 'write', overwrite=args.overwrite)
# Copy the first file's metadata
args.qprint("Reading First File and Establishing Output")
if args.main_header_file:
if args.main_header_file not in args.mvf:
raise RuntimeError("{} not found in files".format(
args.main_header_file))
args.main_header_file = args.mvf.index(args.main_header_file)
else:
args.main_header_file = 0
first_mvf = MultiVariantFile(args.mvf[args.main_header_file], 'read')
concatmvf.copy_header(first_mvf)
# Open each MVF file, read headers to make unified header
transformers = []
mvfmetadata = []
inputfiles = []
for mvfname in args.mvf:
args.qprint("Reading headers from {}".format(mvfname))
# This will create a dictionary of samples{old:new}, contigs{old:new}
args.qprint("Processing Headers and Indexing: {}".format(mvfname))
transformer = MvfTransformer()
mvf = MultiVariantFile(mvfname,
'read',
contigindex=(not args.skip_index))
if args.skip_index:
mvf.read_index_file()
mvf.reset_max_contig()
mvfmetadata.append(mvf.metadata)
for i, sid in enumerate(mvf.get_sample_ids()):
if sid not in concatmvf.get_sample_ids():
new_sindex = concatmvf.max_sample_index + 0
concatmvf.max_sample_index += 1
concatmvf.sample_indices.append(new_sindex)
concatmvf.sample_ids.append(sid)
concatmvf.sample_data[new_sindex] = {}
concatmvf.sample_data[new_sindex]['id'] = sid
concatmvf.sample_id_to_index[sid] = new_sindex
transformer.set_label(i, concatmvf.sample_id_to_index[sid])
for cindex in mvf.contig_indices:
if (mvf.contig_data[cindex]['label']
not in concatmvf.contig_label_to_index):
new_cindex = (mvf.contig_data[cindex]['id']
if mvf.contig_data[cindex]['id']
not in concatmvf.contig_ids else
concatmvf.get_next_contig_index())
concatmvf.contig_data[new_cindex] = (
mvf.contig_data[cindex].copy())
else:
new_cindex = concatmvf.contig_label_to_index[
mvf.contig_data[cindex]['label']]
transformer.set_contig(cindex, new_cindex)
transformers.append(transformer)
inputfiles.append(mvf)
# Write output header
args.qprint("Writing headers to merge output")
concatmvf.reset_max_sample()
concatmvf.notes.append(args.command_string)
concatmvf.write_data(concatmvf.get_header())
# Now loop through each file
blank_entry = '-' * len(concatmvf.sample_indices)
for cons_contig in concatmvf.contig_indices:
contig_merged_entries = {}
args.qprint("Merging Contig Index: {}".format(cons_contig))
for ifile, mvffile in enumerate(inputfiles):
if cons_contig not in transformers[ifile].contigs:
continue
localcontig = transformers[ifile].contigs[cons_contig]
if 'idx' not in mvffile.contig_data[localcontig]:
print("not found")
continue
for _, pos, allelesets in mvffile.itercontigentries(localcontig,
decode=True):
if pos not in contig_merged_entries:
contig_merged_entries[pos] = blank_entry[:]
for j, base in enumerate(allelesets[0]):
xcoord = transformers[ifile].labels_rev[j]
if contig_merged_entries[pos][xcoord] != '-':
if contig_merged_entries[pos][xcoord] == base:
continue
if base in '-X':
continue
raise RuntimeError(
("Merging columns have two different bases: "
"{} {} {}").format(
pos, contig_merged_entries[pos][xcoord],
base))
contig_merged_entries[pos] = (
contig_merged_entries[pos][:xcoord] + base +
contig_merged_entries[pos][xcoord + 1:])
if contig_merged_entries:
concatmvf.write_entries(
((cons_contig, coord, (entry, ))
for coord, entry in sorted(contig_merged_entries.items())),
encoded=False)
args.qprint("Entries written for contig {}: {}".format(
cons_contig, len(contig_merged_entries)))
return ''
def legacy_translate_mvf(args):
"""Main method"""
args.qprint("Running LegacyTranslateMVF")
if args.gff:
args.qprint("Reading and Indexing MVF.")
else:
args.qprint("Reading MVF.")
mvf = MultiVariantFile(args.mvf, 'read', contigindex=bool(args.gff))
if mvf.flavor != 'dna':
raise RuntimeError("MVF must be flavor=dna to translate")
if args.gff:
args.qprint("Processing MVF Index File.")
mvf.read_index_file()
args.qprint("GFF processing start.")
gff = parse_gff_legacy_translate(
args.gff, args,
parent_gene_pattern=args.parent_gene_pattern)
args.qprint("GFF processed.")
outmvf = MultiVariantFile(args.out, 'write', overwrite=args.overwrite)
outmvf.copy_headers_from(mvf)
outmvf.flavor = args.output_data
outmvf.write_data(outmvf.get_header())
args.qprint("Output MVF Established.")
entrybuffer = []
nentry = 0
pos = None
if not args.gff:
args.qprint("No GFF used, translating sequences as pre-aligned in "
"coding frame.")
inputbuffer = []
current_contig = ''
for contigid, pos, allelesets in mvf.iterentries(decode=False):
if current_contig == '':
current_contig = contigid[:]
if contigid == current_contig:
inputbuffer.append((pos, allelesets))
else:
for _, amino_acids, alleles in iter_codons(
inputbuffer, mvf):
if all([x in '-X' for x in amino_acids]):
continue
if args.output_data == 'protein':
entrybuffer.append(
(current_contig, pos, (amino_acids,)))
else:
entrybuffer.append((
current_contig, pos, (
amino_acids, alleles[0],
alleles[1], alleles[2])))
nentry += 1
if nentry == args.line_buffer:
outmvf.write_entries(entrybuffer)
entrybuffer = []
nentry = 0
inputbuffer = [(pos, allelesets)]
current_contig = contigid[:]
if inputbuffer:
for _, amino_acids, alleles in iter_codons(
inputbuffer, outmvf):
if all([x in '-X' for x in amino_acids]):
continue
if args.output_data == 'protein':
entrybuffer.append(
(current_contig, pos, (amino_acids,)))
else:
entrybuffer.append((
current_contig, pos, (
amino_acids, alleles[0],
alleles[1], alleles[2])))
nentry += 1
if nentry == args.line_buffer:
outmvf.write_entries(entrybuffer)
entrybuffer = []
nentry = 0
else:
args.qprint("Indexing GFF gene names.")
# mvfid_to_gffname = outmvf.get_contig_reverse_dict()
for xcontig in outmvf.get_contig_indices():
mvf_entries = {}
xcontiglabel = outmvf.get_contig_labels(indices=xcontig)[0]
xcontigid = outmvf.get_contig_ids(indices=xcontig)[0]
if xcontiglabel not in gff:
if args.verbose:
print(
("No entries in GFF, "
"skipping contig: index:{} id:{} label:{}").format(
xcontig, xcontigid, xcontiglabel))
continue
if not xcontig % 100:
args.qprint("Processing contig: {} {}".format(
xcontigid, xcontiglabel))
for contigid, pos, allelesets in mvf.itercontigentries(
xcontig, decode=False):
mvf_entries[pos] = allelesets[0]
for coords in sorted(gff[xcontiglabel]):
reverse_strand = coords[3] == '-'
alleles = (tuple(mvf_entries.get(x, '-')
for x in coords[2::-1])
if reverse_strand is True
else tuple(mvf_entries.get(x, '-')
for x in coords[0:3]))
if all(len(x) == 1 for x in alleles):
if reverse_strand:
alleles = tuple(
MLIB.complement_bases[x] for x in alleles)
decoded_alleles = alleles
amino_acids = translate_single_codon(''.join(alleles))
else:
if reverse_strand is True:
decoded_alleles = tuple(tuple(MLIB.complement_bases[y]
for y in mvf.decode(x))
for x in alleles)
alleles = tuple(outmvf.encode(''.join(x))
for x in decoded_alleles)
else:
decoded_alleles = tuple(mvf.decode(x) for x in alleles)
amino_acids = tuple(translate_single_codon(''.join(x))
for x in zip(*decoded_alleles))
# print("aminx", amino_acids)
amino_acids = outmvf.encode(''.join(amino_acids))
# if all(x in '-X' for x in amino_acids):
# continue
# print("amino", amino_acids)
# print("translated", amino_acids, alleles)
if args.output_data == 'protein':
entrybuffer.append((xcontig, coords[0], (amino_acids,)))
else:
entrybuffer.append((
xcontigid, coords[0], (
amino_acids, alleles[0], alleles[1], alleles[2])))
nentry += 1
if nentry >= args.line_buffer:
args.qprint("Writing a block of {} entries.".format(
args.line_buffer))
outmvf.write_entries(entrybuffer)
entrybuffer = []
nentry = 0
if entrybuffer:
outmvf.write_entries(entrybuffer)
entrybuffer = []
nentry = 0
return ''
def merge_mvf(args):
"""Main method"""
args.qprint("Running MergeMVF")
if any(fpath.endswith('.gz') for fpath in args.mvf):
print("WARNING! Running MergeMVF with gzipped input files is "
"extremely slow and strongly discouraged.")
concatmvf = MultiVariantFile(args.out, 'write', overwrite=args.overwrite)
# Copy the first file's metadata
args.qprint("Reading First File and Establishing Output")
if args.main_header_file:
if args.main_header_file not in args.mvf:
raise RuntimeError("{} not found in files".format(
args.main_header_file))
else:
args.main_header_file = args.mvf.index(args.main_header_file)
else:
args.main_header_file = 0
first_mvf = MultiVariantFile(args.mvf[args.main_header_file], 'read')
concatmvf.metadata = first_mvf.metadata.copy()
# Open each MVF file, read headers to make unified header
transformers = []
mvfmetadata = []
concatmvf_reverse_contig = dict(
(x['label'], k) for (k, x) in concatmvf.metadata['contigs'].items())
inputfiles = []
for mvfname in args.mvf:
args.qprint("Reading headers from {}".format(mvfname))
# This will create a dictionary of samples{old:new}, contigs{old:new}
args.qprint("Processing Headers and Indexing: {}".format(mvfname))
transformer = MvfTransformer()
mvf = MultiVariantFile(mvfname,
'read',
contigindex=(not args.skip_index))
if args.skip_index:
mvf.read_index_file()
mvf.reset_max_contig_id()
mvfmetadata.append(mvf.metadata)
for i, label in enumerate(mvf.get_sample_labels()):
if label not in concatmvf.get_sample_labels():
concatmvf.metadata['labels'].append(label)
concatmvf.metadata['samples'][
concatmvf.metadata['labels'].index(label)] = {
'label': label
}
# if concatmvf.metadata['labels'].index(label) != i:
transformer.set_label(i, concatmvf.metadata['labels'].index(label))
for contigid, contigdata in iter(mvf.metadata['contigs'].items()):
if contigdata['label'] not in concatmvf_reverse_contig:
newid = (contigid
if contigid not in concatmvf.metadata['contigs'] else
concatmvf.get_next_contig_id())
concatmvf.metadata['contigs'][newid] = contigdata
concatmvf_reverse_contig[contigdata['label']] = newid
else:
newid = concatmvf_reverse_contig[contigdata['label']]
transformer.set_contig(contigid, newid)
transformers.append(transformer)
inputfiles.append(mvf)
# Write output header
args.qprint("Writing headers to merge output")
concatmvf.reset_ncol()
concatmvf.write_data(concatmvf.get_header())
contigs = concatmvf.metadata['contigs']
# Now loop through each file
blank_entry = '-' * len(concatmvf.metadata['samples'])
for current_contig in contigs:
contig_merged_entries = {}
args.qprint("Merging Contig: {}".format(current_contig))
for ifile, mvffile in enumerate(inputfiles):
if current_contig not in transformers[ifile].contigs:
continue
localcontig = transformers[ifile].contigs[current_contig]
for chrom, pos, allelesets in mvffile.itercontigentries(
localcontig, decode=True):
if pos not in contig_merged_entries:
contig_merged_entries[pos] = blank_entry[:]
for j, base in enumerate(allelesets[0]):
xcoord = transformers[ifile].labels_rev[j]
if contig_merged_entries[pos][xcoord] != '-':
if contig_merged_entries[pos][xcoord] == base:
continue
if base == '-' or base == 'X':
continue
raise RuntimeError(
"Merging columns have two different bases: {} {} {}"
.format(pos, contig_merged_entries[pos][xcoord],
base))
contig_merged_entries[pos] = (
contig_merged_entries[pos][:xcoord] + base +
contig_merged_entries[pos][xcoord + 1:])
concatmvf.write_entries(
((current_contig, coord, (entry, ))
for coord, entry in sorted(contig_merged_entries.items())),
encoded=False)
args.qprint("Entries written for contig {}: {}".format(
current_contig, len(contig_merged_entries)))
return ''