def maf2mvf(args):
"""Main method"""
# ESTABLISH MAF
args.qprint("Starting ConvertMAF2MVF")
maf = MultiAlignFile(args)
args.qprint("MAF Established")
# ESTABLISH MVF
mvf = MultiVariantFile(args.out, 'write', overwrite=args.overwrite)
args.qprint("MVF output initialized")
# PROCESS SAMPLE INFO
contig_translate = {1: 1}
samplelabels = [s.split(':')[0] for s in args.sample_tags.split(',')]
args.qprint("Sample tags processed: {}".format(samplelabels))
if args.ref_tag not in samplelabels:
raise IndexError("--ref-tag not in the tags listed in --sample-tags")
samplelabels.remove(args.ref_tag)
samplelabels.insert(0, args.ref_tag)
mvf.sample_ids = samplelabels[:]
mvf.sample_indices = list(range(len(mvf.sample_ids)))
for i, label in enumerate(samplelabels):
mvf.sample_data[i] = {'id': label, 'index': i}
mvf.reset_max_sample()
mvf.metadata['sourceformat'] = maf.metadata['sourceformat']
mvf.metadata.notes.append(args.command_string)
# WRITE MVF HEADER
mvf.write_data(mvf.get_header())
args.qprint("MAF Headers Written")
mvfentries = []
nentry = 0
total_entries = 0
args.qprint("Begin data conversion")
for pos, length, msa in maf:
for sname in samplelabels:
if sname not in msa:
msa[sname] = '-'*length
msa['contig'] = 1
for i in range(length):
mvf_alleles = encode_mvfstring(
''.join(msa[s][i].strip() for s in samplelabels))
if mvf_alleles:
mvfentries.append(
(contig_translate.get(msa['contig']),
pos+i, (mvf_alleles,)))
nentry += 1
if nentry == args.line_buffer:
total_entries += nentry
mvf.write_entries(mvfentries, encoded=True)
args.qprint("{} entries written".format(total_entries))
mvfentries = []
nentry = 0
if mvfentries:
total_entries += nentry
mvf.write_entries(mvfentries)
args.qprint("{} entries written".format(total_entries))
args.qprint("Complete.")
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 ''