def mvf2phy(args):
"""Main method"""
mvf = MultiVariantFile(args.mvf, 'read')
if (mvf.flavor in ("dna", "rna") and args.output_data == "prot") or (
mvf.flavor == "prot" and args.output_data in ("dna", "rna")):
raise RuntimeError(
"--outdput-data {} incompatiable with '{}' flavor mvf".format(
args.output_data, mvf.flavor))
max_region_coord = dict((x, None) for x in mvf.get_contig_ids())
if args.regions is not None:
_, max_region_coord, _ = parse_regions_arg(args.regions,
mvf.get_contig_ids())
if args.sample_indices is not None:
sample_indices = [int(x) for x in args.sample_indices[0].split(",")]
elif args.sample_labels is not None:
sample_indices = mvf.get_sample_indices(
ids=args.sample_labels[0].split(","))
else:
sample_indices = mvf.get_sample_indices()
sample_labels = mvf.get_sample_ids(indices=sample_indices)
skipcontig = ''
tmp_files = dict((fn,
open("{}-{}.tmp".format(fn, randint(1000000, 9999999)),
'w+', args.buffer)) for fn in sample_labels)
labelwritten = dict.fromkeys(sample_labels, False)
current_contig_id = None
current_contig_start = 1
current_contig_end = 1
if args.partition is True:
partprefix = "PROT" if args.output_data == "prot" else "DNA"
partitionfile = open("{}.part".format(args.out), 'w')
for contig, _, allelesets in mvf.iterentries(
contig_ids=(mvf.get_contig_ids()
if args.regions is None else max_region_coord[:]),
decode=True):
if contig == skipcontig:
continue
if contig not in max_region_coord:
skipcontig = contig[:]
continue
if current_contig_id is None:
current_contig_id = contig[:]
elif contig != current_contig_id:
if args.partition is True:
if current_contig_end > current_contig_start:
partitionfile.write("{}, {} = {}-{}\n".format(
partprefix,
mvf.get_contig_labels(ids=current_contig_id),
current_contig_start, current_contig_end - 1))
current_contig_id = contig[:]
# reset start as one position after end of last
current_contig_start = current_contig_end
current_contig_end = current_contig_end + 1
for col, label in zip(sample_indices, sample_labels):
if not labelwritten[label]:
if args.label_type == 'long':
tmp_files[label].write("{}{}".format(
label[:100], " " * (100 - len(label[:100]))))
elif args.label_type == 'short':
tmp_files[label].write("{}{}".format(
label[:20], " " * (20 - len(label[:20]))))
labelwritten[label] = True
if mvf.flavor == 'dna':
tmp_files[label].write(allelesets[0][col] == 'X' and 'N'
or allelesets[0][col])
if label == sample_labels[0]:
current_contig_end += 1
elif ((mvf.flavor == 'codon' and args.output_data == 'prot')
or (mvf.flavor == 'prot')):
tmp_files[label].write(allelesets[0][col])
if label == sample_labels[0]:
current_contig_end += 1
elif mvf.flavor == 'codon':
codon = [
"N" if allelesets[x][col] == 'X' else allelesets[x][col]
for x in (1, 2, 3)
]
tmp_files[label].write(''.join(codon))
if label == sample_labels[0]:
current_contig_end += 3
first_file = True
totalseqlen = 0
with open(args.out, 'w') as outfile:
for filehandler in tmp_files.values():
# read first file to establish sequence length for phylip header
if first_file is True:
filehandler.seek(0, 0)
buff = filehandler.read(args.buffer)
while buff != '':
if " " in buff:
totalseqlen += len(buff.strip().split(" ")[-1])
else:
totalseqlen += len(buff.strip())
buff = filehandler.read(args.buffer)
outfile.write("{} {}\n".format(len(sample_labels),
totalseqlen))
first_file = False
filehandler.seek(0, 0)
buff = filehandler.read(args.buffer)
while buff != '':
if first_file is True:
outfile.write("{} {}\n".format(len(sample_labels),
len(buff.split()[1])))
first_file = False
outfile.write(buff)
buff = filehandler.read(args.buffer)
outfile.write("\n")
filehandler.close()
os.remove(os.path.join(args.temp_dir, filehandler.name))
if args.partition is True:
if current_contig_end > current_contig_start:
partitionfile.write("{}, {} = {}-{}\n".format(
partprefix, mvf.get_contig_labels(ids=current_contig_id),
current_contig_start, current_contig_end - 1))
partitionfile.close()
return ''
def vcf2mvf(args=None):
"""Main method for vcf2mvf"""
sepchars = dict([("TAB", "\t"), ("SPACE", " "), ("DBLSPACE", " "),
("COMMA", ","), ("MIXED", None)])
args.fieldsep = sepchars[args.field_sep]
# ESTABLISH VCF
args.qprint("Opening input VCF: {}".format(args.vcf))
vcf = VariantCallFile(args.vcf, indexcontigs=(not args.no_autoindex))
# ESTABLISH MVF
args.qprint("Establishing output MVF: {}".format(args.out))
mvf = MultiVariantFile(args.out, 'write', overwrite=args.overwrite)
mvf.notes.append(args.command_string)
mvf.metadata['mvfversion'] = args.versionx
# PROCESS CONTIG INFO
args.qprint("Processing VCF headers.")
vcfcontigs = vcf.metadata['contigs'].copy()
args.qprint("{} contigs found.".format(len(vcfcontigs)))
contig_translate = {}
if args.contig_ids:
for cid, cvcf, cmvf in (x.split(';') for x in args.contig_ids):
try:
cid = int(cid)
except ValueError:
pass
assert cvcf in [vcfcontigs[x]['label'] for x in vcfcontigs]
for vid in vcfcontigs:
if vcfcontigs[vid]['label'] == cvcf:
contig_translate[cvcf] = [cid, cmvf]
if cid in mvf.metadata['contigs']:
raise RuntimeError(
'Contig id {} is not unique'.format(cid))
mvf.metadata['contigs'][cid] = vcfcontigs[vid].copy()
if cmvf in mvf.get_contig_labels():
raise RuntimeError(
'Contig label {} is not unique'.format(cmvf))
mvf.metadata['contigs'][cid]['label'] = cmvf[:]
mvf.reset_max_contig()
mvf.max_contig_index -= 1
args.qprint("Processing contigs.")
static_contig_ids = list(mvf.get_contig_ids())
for vcid in vcfcontigs:
vlabel = vcfcontigs[vcid]['label']
if vlabel not in static_contig_ids:
newindex = mvf.get_next_contig_index()
if ((is_int(vlabel) or len(vlabel) < 3)
and vlabel not in static_contig_ids):
newid = vlabel[:]
else:
newid = str(newindex)
mvf.contig_indices.append(newindex)
mvf.contig_ids.append(newid)
mvf.contig_data[newindex] = vcfcontigs[vcid].copy()
static_contig_ids.append(newid)
contig_translate[vlabel] = [newindex, vlabel]
mvf.reset_max_contig()
new_contigs = [(x, mvf.contig_data[x]['label'])
for x in mvf.contig_indices]
if args.skip_contig_label_check is False:
args.qprint("Checking contigs for label/id overlap errors.")
xids = [x[0] for x in new_contigs]
xlabels = [x[1] for x in new_contigs]
xintersect = set(xids).intersection(xlabels)
if xintersect:
for i, (newid, newlabel) in enumerate(new_contigs):
if i % 100 == 0:
args.qprint("{} contigs processed".format(i))
if newid in xlabels[:i] or newid in xlabels[i + 1:]:
# if newid in xlabels:
# if xlabels.index(newid) != i:
raise RuntimeError("Error contig id {} is the same as"
" the label for another contig"
" ({})".format(newid,
xlabels.index(newid)))
if newlabel in xids[:i] or newlabel in xids[i + 1:]:
# if newlabel in xids:
# if xids.index(newlabel) != i:
raise RuntimeError("Error contig label {} is the same"
"as the id for another contig"
"({})".format(newlabel,
xids.index(newlabel)))
# PROCESS SAMPLE INFO
args.qprint("Processing samples.")
samplelabels = [args.ref_label] + vcf.metadata['samples'][:]
if args.alleles_from:
args.alleles_from = args.alleles_from.split(':')
samplelabels += args.alleles_from
if args.sample_replace:
newsample = [
x.split(':') if ':' in tuple(x) else tuple([x, x])
for x in args.sample_replace
]
unmatched = list(enumerate(samplelabels))
for old, new in newsample:
labelmatched = False
for j, (i, name) in enumerate(unmatched):
if old in name:
samplelabels[i] = new
labelmatched = j
break
if labelmatched is not False:
del unmatched[labelmatched]
mvf.sample_indices = list(range(len(samplelabels)))
mvf.sample_ids = samplelabels[:]
for i, label in enumerate(samplelabels):
mvf.sample_data[i] = {'id': label}
mvf.metadata['ncol'] = len(mvf.sample_ids)
mvf.max_sample_index = len(mvf.sample_ids)
mvf.metadata['sourceformat'] = vcf.metadata['sourceformat']
# WRITE MVF HEADER
mvf.write_data(mvf.get_header())
mvfentries = []
nentry = 0
args.qprint("Processing VCF entries.")
for vcfrecord in vcf.iterentries(args):
mvfstring = ''.join(vcfrecord['genotypes'])
if args.filter_nonref_empty is True:
if all(x in 'Xx-?' for x in mvfstring[1:]):
continue
mvf_alleles = encode_mvfstring(mvfstring)
if args.out_flavor in ('dnaqual', ):
qual_alleles = encode_mvfstring(''.join(vcfrecord['qscores']))
if mvf_alleles:
mvfentries.append(
(contig_translate.get(vcfrecord['contig'])[0],
vcfrecord['coord'],
((mvf_alleles,
qual_alleles) if args.out_flavor in ('dnaqual', ) else
(mvf_alleles, ))))
nentry += 1
if nentry == args.line_buffer:
mvf.write_entries(mvfentries, encoded=True)
mvfentries = []
nentry = 0
if mvfentries:
mvf.write_entries(mvfentries)
mvfentries = []
return ''
def mvf2fastagene(args):
"""Main method"""
args.qprint("Indexing MVF")
mvf = MultiVariantFile(args.mvf, 'read', contigindex=True)
if (mvf.flavor in ("dna", "rna") and args.output_data == "prot") or (
mvf.flavor == "prot" and args.output_data in ("dna", "rna")):
raise RuntimeError(
"--output-data {} incompatiable with '{}' flavor mvf".format(
args.output_data, mvf.flavor))
if args.output_data is None:
raise RuntimeError("--output-data required")
sample_labels = mvf.get_sample_ids()
if args.sample_indices is not None:
sample_indices = [int(x) for x in args.sample_indices[0].split(",")]
elif args.sample_labels is not None:
sample_indices = mvf.get_sample_indices(
ids=args.sample_labels[0].split(","))
else:
sample_indices = mvf.get_sample_indices()
args.qprint("Beginning Entries.")
if not os.path.exists(args.output_dir):
args.qprint("Output Directory Created: {}".format(args.output_dir))
os.mkdir(args.output_dir)
else:
args.qprint("Output Directory Exists Already: {}".format(
args.output_dir))
write_buffer = {}
for targetcontig in mvf.get_contig_indices():
contiglabel = mvf.get_contig_labels(indices=targetcontig)[0]
args.qprint("Reading Contig {}: {}".format(targetcontig, contiglabel))
write_buffer = dict((x, []) for x in sample_labels)
data_in_buffer = False
for _, _, allelesets in mvf.itercontigentries(targetcontig,
decode=True):
for col, label in zip(sample_indices, sample_labels):
if mvf.flavor == 'dna':
write_buffer[label].append('N' if allelesets[0][col] ==
'X' else allelesets[0][col])
data_in_buffer = True
elif mvf.flavor in ('codon', 'prot') and (args.output_data
== 'prot'):
write_buffer[label].append(allelesets[0][col])
data_in_buffer = True
elif mvf.flavor == 'codon' and args.output_data == 'dna':
if args.choose_allele == 'random1':
codon = [
'N' if allelesets[x][col] == 'X' else
(MLIB.randomnuc(allelesets[x][col]) if
(allelesets[x][col]
in MLIB.validchars['dnaambig23']) else
allelesets[x][col]) for x in (1, 2, 3)
]
else:
codon = [
'N' if allelesets[x][col] == 'X' else
allelesets[x][col] for x in (1, 2, 3)
]
write_buffer[label].append(''.join(codon))
data_in_buffer = True
if data_in_buffer:
args.qprint("Writing Align")
with open(os.path.join(args.output_dir, contiglabel + ".fa"),
'w') as outfile:
for label in write_buffer:
if (mvf.flavor == 'codon'
and args.output_data in ('dna', 'prot')):
if ((mvf.contig_data[targetcontig].get('strand', '+')
== '-') and (args.ignore_strand is False)):
entryseq = ''.join(write_buffer[label][::-1])
else:
entryseq = ''.join(write_buffer[label])
else:
entryseq = ''.join(write_buffer[label])
outfile.write(">{}\n{}\n".format(label, entryseq))
outfile.write("\b")
return ''
def calc_group_unique_allele_window(args):
"""Count the number of and relative rate of uniquely held alleles
spatially along chromosomes (i.e. Lineage-specific rates)"""
data = {}
mvf = MultiVariantFile(args.mvf, 'read')
if mvf.flavor != 'codon':
raise RuntimeError(
"\n=====================\nERROR: MVF is not codon flavor!")
annotations = {}
coordinates = {}
labels = mvf.get_sample_labels()[:]
ncol = len(labels)
current_contig = None
current_position = 0
counts = Counter()
totals = Counter()
args.start_contig = (args.start_contig
if args.start_contig is not None else 0)
args.end_contig = (args.end_contig
if args.end_contig is not None else 100000000000)
if args.output_align is True:
outputalign = []
if args.gff is not None:
annotations, coordinates = (parse_gff_analysis(args.gff))
if args.allele_groups is not None:
args.allele_groups = procarg_allelegroups(args.allele_groups, mvf)
if args.species_groups is None:
args.species_groups = args.allele_groups
else:
args.species_groups = procarg_speciesgroups(args.species_groups, mvf)
fieldtags = [
'likelihood', 'bgdnds0', 'bgdnds1', 'bgdnds2a', 'bgdnds2b', 'fgdnds0',
'fgdnds1', 'fgdnds2a', 'fgdnds2b', 'dndstree', 'errorstate'
]
if args.branch_lrt is not None:
with open(args.branch_lrt, 'w') as branchlrt:
genealign = []
branchlrt.write(
"\t".join(['contig', 'ntaxa', 'alignlength', 'lrtscore'] +
["null.{}".format(x) for x in fieldtags] +
["test.{}".format(x)
for x in fieldtags] + ['tree']) + "\n")
groups = args.allele_groups.values()
if args.species_groups is not None:
speciesgroups = args.species_groups.values()
allsets = set([])
for group in groups:
allsets.update(group)
allsets = list(sorted(allsets))
speciesnames = args.species_groups.keys()
speciesrev = {}
if args.species_groups is not None:
for species in args.species_groups:
speciesrev.update([(x, species)
for x in args.species_groups[species]])
if args.mincoverage is not None:
if args.mincoverage < len(groups) * 2:
raise RuntimeError("""
Error: GroupUniqueAlleleWindow:
--mincoverage cannot be lower than the twice the number
of specified groups in --allele-groups
""")
genealign = []
for contig, pos, allelesets in mvf:
if not current_contig:
current_contig = contig[:]
if contig != current_contig or (args.windowsize > 0 and pos >
current_position + args.windowsize):
xkey = (
current_contig,
current_position,
)
data[xkey] = counts.copy()
data[xkey].update([
('contig', (mvf.get_contig_labels(ids=current_contig)
if args.use_labels is True else current_contig)),
('position', current_position),
('nonsynyonymous_changes',
counts.get('nonsynonymous_changes', 0) or 0),
('synyonymous_changes', counts.get('synonymous_changes', 0)
or 0)
])
data[xkey].update([
('ns_ratio',
(float(data[xkey].get('nonsynonymous_changes', 0)) /
(data[xkey].get('synonymous_changes', 1.0)))),
('annotation', annotations.get(data[xkey]['contig'], '.')),
('coordinates', coordinates.get(data[xkey]['contig'], '.'))
])
if genealign:
if (args.end_contig >= int(current_contig)) and (
args.start_contig <= int(current_contig)):
(pamlnull, pamltest, tree) = paml_branchsite(
genealign,
labels[:],
species=speciesnames,
speciesrev=speciesrev,
codemlpath=args.codeml_path,
raxmlpath=args.raxml_path,
pamltmp=args.paml_tmp,
target=args.target,
targetspec=args.num_target_species,
allsampletrees=args.all_sample_trees,
outgroup=args.outgroup)
lrtscore = -1
if (pamlnull.get('likelihood', -1) != -1
and pamltest.get('likelihood', -1) != -1):
lrtscore = 2 * (pamltest['likelihood'] -
pamlnull['likelihood'])
with open(args.branch_lrt, 'a') as branchlrt:
branchlrt.write("\t".join([
str(x) for x in [
data[xkey]['contig'],
len(genealign),
len(genealign[0]) * 3, lrtscore
] + [pamlnull.get(y, -1) for y in fieldtags] +
[pamltest.get(y, -1)
for y in fieldtags] + [str(tree).rstrip()]
]) + "\n")
genealign = None
totals.add('genes_total')
if counts.get('total_codons', 0) > 0:
totals.add('genes_tested')
if counts.get('total_nsyn_codons', 0) > 0:
totals.add('genes_with_nsyn')
if contig != current_contig:
current_contig = contig[:]
current_position = 0
elif args.windowsize > 0:
current_position += args.windowsize
counts = Counter()
proteins = allelesets[0]
codons = allelesets[1:4]
if len(proteins) == 1 and all(len(x) == 1 for x in codons):
if proteins == '*' or ''.join(codons) in MLIB.stop_codons:
continue
counts.add('total_codons')
totals.add('total_codons')
if args.output_align is True:
if not outputalign:
outputalign = [[''.join(codons)]
for x in range(mvf.metadata['ncol'])]
else:
for ialign, xalign in enumerate(outputalign):
xalign.append(''.join(codons))
if args.branch_lrt is not None:
if not genealign:
genealign = [[''.join(codons)] for x in range(ncol)]
else:
for ialign in range(len(genealign)):
genealign[ialign].append(''.join(codons))
continue
if len(proteins) > 1:
if allelesets[0][1] == '+':
continue
proteins = mvf.decode(proteins)
if args.mincoverage is not None:
if sum([int(x not in 'X-')
for x in proteins]) < (args.mincoverage):
continue
species_groups = [[proteins[i] for i in x if proteins[i] not in '-X']
for x in speciesgroups]
if any(len(x) == 0 for x in species_groups):
continue
xcodons = [mvf.decode(x) for x in codons]
codons = [''.join(x) for x in zip(*xcodons)]
if any(codons[x] in MLIB.stop_codons for x in allsets):
continue
if any(
any(x != species_groups[0][0] for x in y)
for y in species_groups):
totals.add('total_nsyn_codons')
counts.add('total_nsyn_codons')
totals.add('total_codons')
totals.add('tested_codons')
counts.add('total_codons')
totals.add('variable_codons',
val=int(
sum([int(len(set(x) - set('X-')) > 1)
for x in xcodons]) > 0))
if args.output_align is not None:
if not outputalign:
outputalign = [[x] for x in codons]
else:
for ialign in range(len(outputalign)):
outputalign[ialign].append(codons[ialign])
if args.branch_lrt is not None:
if not genealign:
genealign = [[x] for x in codons]
else:
for ialign in range(len(codons)):
genealign[ialign].append(codons[ialign])
nonsyn_change = False
synon_change = False
codon_groups = [
set([
codons[i] for i in x
if '-' not in codons[i] and 'X' not in codons[i]
]) for x in groups
]
protein_groups = None
for i in range(len(codon_groups)):
if any(base in codon for base in 'RYWKMS'
for codon in codon_groups[i]):
codon_groups[i] = hapgroup(codon_groups[i])
if all(
grp1.isdisjoint(grp0)
for grp0, grp1 in combinations(codon_groups, 2)):
protein_groups = [
set([
MLIB.codon_tables['full'][''.join(x)]
for x in codon_groups[i]
]) for i in range(len(codon_groups))
]
if all(
grp1.isdisjoint(grp0)
for grp0, grp1 in combinations(protein_groups, 2)):
nonsyn_change = True
elif all(grp1 == grp0
for grp0, grp1 in combinations(protein_groups, 2)):
synon_change = True
if nonsyn_change:
if args.verbose is True:
print('NON', contig, pos, allelesets,
codon_groups, protein_groups, groups,
mvf.get_contig_labels(ids=contig))
counts.add('nonsynonymous_changes')
totals.add('nonsynonymous_changes')
elif synon_change:
if args.verbose is True:
print('SYN', contig, pos, allelesets,
codon_groups, protein_groups, groups,
mvf.get_contig_labels(ids=contig))
counts.add('synonymous_changes')
totals.add('synonymous_changes')
args.totals = totals
# WRITE OUTPUT
headers = [
"contig", "position", "nonsynonymous_changes", "synonymous_changes",
"ns_ratio", "nonsynonymous_total", "synonymous_total", "pvalue",
"total_codons", "annotation", "coordinates"
]
if args.windowsize == -1:
headers.remove('position')
if args.chi_test is None:
headers.remove('pvalue')
outfile = OutputFile(path=args.out, headers=headers)
sorted_entries = sorted(
[(data[k]['ns_ratio'], k)
for k in data if data[k].get('nonsynonymous_changes', 0) > 0],
reverse=True)
for _, k in sorted_entries:
outfile.write_entry(data[k])
with open(args.out + '.total', 'w') as totalfile:
for entry in args.totals.iter_sorted():
totalfile.write(entry)
if args.output_align is not None:
with open(args.output_align, 'w') as alignfile:
alignfile.write("\n".join([
">{}\n{}".format(mvf.metadata['labels'][i],
''.join(outputalign[i]))
for i in range(len(outputalign))
]))
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 infer_window_tree(args):
"""Main method"""
# ESTABLISH FILE OBJECTS
mvf = MultiVariantFile(args.mvf, 'read')
# Set up contig ids
if args.contig_ids is not None:
contig_ids = args.contig_ids[0].split(",")
elif args.contig_labels is not None:
contig_ids = mvf.get_contig_ids(
labels=args.contig_labels[0].split(","))
else:
contig_ids = mvf.get_contig_ids()
treefile = OutputFile(
args.out,
headers=[
'contig',
'windowstart',
'windowsize',
'tree',
'topology',
'topoid',
# 'templabels', ### USED FOR DEBUGGING ###
'alignlength',
'aligndepth',
'status'
])
topofile = OutputFile(args.out + '.counts',
headers=['rank', 'topology', 'count'])
if args.sample_indices is not None:
sample_indices = [int(x) for x in args.sample_indices[0].split(",")]
elif args.sample_labels is not None:
sample_indices = mvf.get_sample_indices(
labels=args.sample_labels[0].split(","))
else:
sample_indices = mvf.get_sample_indices()
if not os.path.exists(args.temp_dir):
os.mkdir(args.temp_dir)
os.chdir(args.temp_dir)
# SETUP PARAMS
main_labels = mvf.get_sample_labels(sample_indices)
if args.choose_allele in ['randomboth', 'majorminor']:
main_labels = [label + x for x in ['a', 'b'] for label in main_labels]
params = {
'outgroups':
args.raxml_outgroups or [],
'rootwith':
(args.root_with.split(',') if args.root_with is not None else None),
'minsites':
args.min_sites,
'minseqcoverage':
args.min_seq_coverage,
'mindepth':
args.min_depth,
'raxmlpath':
args.raxml_path,
'raxmlopts':
args.raxml_opts,
'duplicateseq':
args.duplicate_seq,
'model':
args.raxml_model,
'bootstrap':
args.bootstrap,
'windowsize':
args.windowsize,
'chooseallele':
args.choose_allele,
'tempdir':
args.temp_dir,
'tempprefix':
args.temp_prefix
}
# WINDOW START INTERATION
verify_raxml(params)
current_contig = ''
current_position = 0
window_data = None
skip_contig = False
topo_ids = {}
topo_counts = {}
for contig, pos, allelesets in mvf.iterentries(contigs=contig_ids,
subset=sample_indices,
quiet=args.quiet,
no_invariant=False,
no_ambig=False,
no_gap=False,
decode=True):
if current_contig == contig:
if skip_contig is True:
continue
if not same_window((current_contig, current_position),
(contig, pos), args.windowsize):
skip_contig = False
if window_data is not None:
entry = window_data.maketree_raxml(params)
if entry['status'] != 'ok':
if args.output_empty:
treefile.write_entry(entry)
if args.windowsize != -1:
skip_contig = True
else:
topo = entry["topology"]
topo_counts[topo] = topo_counts.get(topo, 0) + 1
if topo not in topo_ids:
topo_ids[topo] = (topo_ids
and max(topo_ids.values()) + 1 or 0)
entry["topoid"] = topo_ids[topo]
treefile.write_entry(entry)
current_position = (current_position + args.windowsize if
(contig == current_contig
and args.windowsize > 0) else 0)
current_contig = contig[:]
window_data = None
window_data = WindowData(
window_params={
'contigname': (mvf.get_contig_labels(
ids=current_contig) if args.output_contig_labels
is not None else current_contig[:]),
"windowstart": (
'-1' if args.windowsize == -1 else current_position +
0),
"windowsize":
args.windowsize,
"labels":
main_labels[:]
})
# ADD ALLELES
if mvf.flavor == 'dna':
if args.choose_allele != 'none':
allelesets[0] = hapsplit(allelesets[0], args.choose_allele)
window_data.append_alleles(allelesets[0], mindepth=args.min_depth)
# LAST LOOP
if window_data:
entry = window_data.maketree_raxml(params)
if entry['status'] != 'ok':
if args.output_empty:
treefile.write_entry(entry)
else:
topo = entry["topology"]
topo_counts[topo] = topo_counts.get(topo, 0) + 1
if topo not in topo_ids:
topo_ids[topo] = (max(topo_ids.values()) +
1 if topo_ids else 0)
entry["topoid"] = topo_ids[topo]
treefile.write_entry(entry)
window_data = None
# END WINDOW ITERATION
topo_list = sorted([(v, k) for k, v in topo_counts.items()], reverse=True)
for rank, [value, topo] in enumerate(topo_list):
topofile.write_entry({'rank': rank, 'count': value, 'topology': topo})
return ''
def infer_window_tree(args):
"""Main method"""
args.qprint("Running InferTree")
# ESTABLISH FILE OBJECTS
mvf = MultiVariantFile(args.mvf, 'read')
args.qprint("Read MVF File: {}".format(args.mvf))
# Set up contig ids
if args.contig_ids is not None:
contig_ids = args.contig_ids[0].split(",")
elif args.contig_labels is not None:
contig_ids = mvf.get_contig_ids(
labels=args.contig_labels[0].split(","))
else:
contig_ids = mvf.get_contig_ids()
treefile = OutputFile(
args.out,
headers=['contig', 'windowstart', 'windowsize', 'tree',
'topology', 'topoid',
# 'templabels', ### USED FOR DEBUGGING ###
'alignlength', 'aligndepth', 'status'])
topofile = OutputFile(args.out + '.counts',
headers=['rank', 'topology', 'count'])
if args.sample_indices is not None:
sample_indices = [int(x) for x in
args.sample_indices[0].split(",")]
elif args.sample_labels is not None:
sample_indices = mvf.get_sample_indices(
ids=args.sample_labels[0].split(","))
else:
sample_indices = mvf.get_sample_indices()
if not os.path.exists(args.temp_dir):
os.mkdir(args.temp_dir)
os.chdir(args.temp_dir)
# SETUP PARAMS
main_labels = mvf.get_sample_ids(sample_indices)
if args.choose_allele in ['randomboth', 'majorminor']:
main_labels = [label + x for x in ['a', 'b'] for label in main_labels]
params = {
'bootstrap': args.bootstrap,
'chooseallele': args.choose_allele,
'collapse_polytomies': args.collapse_polytomies,
'duplicateseq': args.duplicate_seq,
'engine': args.engine,
'engine_path': args.engine_path,
'engine_opts': args.engine_opts,
'mindepth': args.min_depth,
'minseqcoverage': args.min_seq_coverage,
'minsites': args.min_sites,
'model': args.model,
'outgroups': (args.raxml_outgroups
if args.raxml_outgroups is not None
else None),
'rootwith': (args.root_with.split(',')
if args.root_with is not None
else []),
'tempdir': args.temp_dir,
'tempprefix': args.temp_prefix,
'windowsize': args.windowsize,
}
# DEFAULT MODEL
if params['model'] is None:
if params['engine'] == 'raxml':
params['model'] = 'GTRGAMMA'
elif params['engine'] == 'raxml-ng':
params['model'] = "GTR+G"
# WINDOW START INTERATION
verify_raxml(params)
args.qprint("RAxML Found.")
current_contig = None
current_position = 0
window_data = None
# skip_contig = False
topo_ids = {}
topo_counts = {}
args.qprint("Prcocessing Records")
windowsizename = "window size={}".format(args.windowsize)
if windowsizename == "window size=-1":
windowsizename = "whole contig"
elif windowsizename == "window size=0":
windowsizename = "whole genome"
window_data = WindowData(window_params={
'contigname': 'all',
"windowstart": 0,
"windowsize": 0,
"labels": main_labels[:]})
for contig, pos, allelesets in mvf.iterentries(
contig_ids=contig_ids, subset=sample_indices,
no_invariant=False, no_ambig=False, no_gap=False, decode=True):
# if current_contig == contig:
# if skip_contig is True:
# args.qprint("Skipping contig: {}".format(current_contig))
# continue
if not same_window((current_contig, current_position),
(contig, pos), args.windowsize):
# skip_contig = False
if window_data is not None:
args.qprint(("Making tree for {} "
"at contig {} position {}").format(
windowsizename,
current_contig,
current_position))
entry = window_data.maketree_raxml(params)
if entry['status'] != 'ok':
if args.output_empty:
treefile.write_entry(entry)
# if args.windowsize != -1:
# skip_contig = True
args.qprint(
"TREE REJECTED with error code: {} ({})".format(
entry['status'], entry.get('comment', "None")))
else:
args.qprint("Tree completed.")
topo = entry["topology"]
topo_counts[topo] = topo_counts.get(topo, 0) + 1
if topo not in topo_ids:
topo_ids[topo] = (max(topo_ids.values()) + 1
if topo_ids else 0)
entry["topoid"] = topo_ids[topo]
treefile.write_entry(entry)
current_position = current_position + args.windowsize if (
contig == current_contig and args.windowsize > 0) else 0
current_contig = contig[:]
window_data = None
window_data = WindowData(window_params={
'contigname': (mvf.get_contig_labels(ids=current_contig) if
args.output_contig_labels is not None else
current_contig[:]),
"windowstart": ('-1' if args.windowsize == -1
else current_position + 0),
"windowsize": args.windowsize,
"labels": main_labels[:]})
# ADD ALLELES
if mvf.flavor == 'dna':
if args.choose_allele != 'none':
allelesets[0] = hapsplit(allelesets[0], args.choose_allele)
window_data.append_alleles(allelesets[0], mindepth=args.min_depth)
elif mvf.flavor == 'codon':
for i in (1, 2, 3):
if args.choose_allele != 'none':
allelesets[i] = hapsplit(allelesets[i], args.choose_allele)
window_data.append_alleles(allelesets[i], mindepth=args.min_depth)
# LAST LOOP
if window_data:
entry = window_data.maketree_raxml(params)
if entry['status'] != 'ok':
if args.output_empty:
treefile.write_entry(entry)
else:
topo = entry["topology"]
topo_counts[topo] = topo_counts.get(topo, 0) + 1
if topo not in topo_ids:
topo_ids[topo] = (
max(topo_ids.values()) + 1 if topo_ids else 0)
entry["topoid"] = topo_ids[topo]
treefile.write_entry(entry)
window_data = None
# END WINDOW ITERATION
topo_list = sorted([(v, k) for k, v in topo_counts.items()],
reverse=True)
for rank, [value, topo] in enumerate(topo_list):
topofile.write_entry({'rank': rank, 'count': value, 'topology': topo})
return ''
def vcf2mvf(args=None):
"""Main method for vcf2mvf"""
sepchars = dict([("TAB", "\t"), ("SPACE", " "), ("DBLSPACE", " "),
("COMMA", ","), ("MIXED", None)])
args.fieldsep = sepchars[args.field_sep]
# ESTABLISH VCF
vcf = VariantCallFile(args.vcf, indexcontigs=(not args.no_autoindex))
# ESTABLISH MVF
mvf = MultiVariantFile(args.out, 'write', overwrite=args.overwrite)
# PROCESS CONTIG INFO
vcfcontigs = vcf.metadata['contigs'].copy()
contig_translate = {}
if args.contig_ids:
for cid, cvcf, cmvf in (x.split(';') for x in args.contig_ids):
try:
cid = int(cid)
except ValueError:
pass
assert cvcf in [vcfcontigs[x]['label'] for x in vcfcontigs]
for vid in vcfcontigs:
if vcfcontigs[vid]['label'] == cvcf:
contig_translate[cvcf] = [cid, cmvf]
if cid in mvf.metadata['contigs']:
raise RuntimeError(
'Contig id {} is not unique'.format(cid))
mvf.metadata['contigs'][cid] = vcfcontigs[vid].copy()
if cmvf in mvf.get_contig_labels():
raise RuntimeError(
'Contig label {} is not unique'.format(cmvf))
mvf.metadata['contigs'][cid]['label'] = cmvf[:]
mvf.reset_max_contig_id()
for vcid in vcfcontigs:
vlabel = vcfcontigs[vcid]['label']
if vlabel not in mvf.get_contig_labels():
if ((is_int(vlabel) or len(vlabel) < 3)
and vlabel not in mvf.get_contig_ids()):
newid = vlabel[:]
else:
newid = mvf.get_next_contig_id()
mvf.metadata['contigs'][newid] = vcfcontigs[vcid].copy()
contig_translate[vlabel] = [newid, vlabel]
mvf.reset_max_contig_id()
new_contigs = [(x, mvf.metadata['contigs'][x]['label'])
for x in mvf.metadata['contigs']]
for i, (newid, newlabel) in enumerate(new_contigs):
for j, (xid, xlabel) in enumerate(new_contigs):
if i == j:
continue
if newid == xlabel:
raise RuntimeError("Error contig id {} is the same as"
" the label for another contig"
" ({} {})".format(newid, xid, xlabel))
if newlabel == xid:
raise RuntimeError("Error contig label {} is the same"
"as the id for another contig"
"({} {})".format(newlabel, xid, xlabel))
# PROCESS SAMPLE INFO
samplelabels = [args.ref_label] + vcf.metadata['samples'][:]
if args.alleles_from:
args.alleles_from = args.alleles_from.split(':')
samplelabels += args.alleles_from
if args.sample_replace:
newsample = [
x.split(':') if ':' in tuple(x) else tuple([x, x])
for x in args.sample_replace
]
unmatched = [x for x in enumerate(samplelabels)]
for old, new in newsample:
labelmatched = False
for j, (i, name) in enumerate(unmatched):
if old in name:
samplelabels[i] = new
labelmatched = j
break
if labelmatched is not False:
del unmatched[labelmatched]
mvf.metadata['labels'] = samplelabels[:]
for i, label in enumerate(samplelabels):
mvf.metadata['samples'][i] = {'label': label}
mvf.metadata['ncol'] = len(mvf.metadata['labels'])
mvf.metadata['sourceformat'] = vcf.metadata['sourceformat']
# WRITE MVF HEADER
mvf.write_data(mvf.get_header())
mvfentries = []
nentry = 0
for vcfrecord in vcf.iterentries(args):
# try:
mvf_alleles = encode_mvfstring(''.join(vcfrecord['genotypes']))
if args.out_flavor in ('dnaqual', ):
qual_alleles = encode_mvfstring(''.join(vcfrecord['qscores']))
if mvf_alleles:
mvfentries.append(
(contig_translate.get(vcfrecord['contig'])[0],
vcfrecord['coord'],
((mvf_alleles,
qual_alleles) if args.out_flavor in ('dnaqual', ) else
(mvf_alleles, ))))
nentry += 1
if nentry == args.line_buffer:
mvf.write_entries(mvfentries, encoded=True)
mvfentries = []
nentry = 0
# except Exception as exception:
if mvfentries:
mvf.write_entries(mvfentries)
mvfentries = []
return ''
