def create_mmseqs_taxonomy(args):
if args.action == 'accessions':
for i, header in enumerate(split_file(args.infile, '\x00')):
accessions = [
x.split(' ')[0].split('.')[0] for x in header.split('\x01')
]
for a in accessions:
print(a)
sys.exit()
tax_db = ncbitax.TaxonomyDb(tax_dir=args.tax_db,
load_names=True,
load_nodes=True)
accession_lookup = {}
start_time = time.time()
log.info('Loading accession to taxid mapping')
if args.genomic_joined:
with ioutil.compressed_open(args.genomic_joined, 'rt') as f:
for line in f:
parts = line.split('\t')
accession = parts[0]
taxid = parts[1]
taxid = int(taxid)
accession_lookup[base_accession] = taxid
elif args.accession2taxid:
for fn in args.accession2taxid:
with ioutil.compressed_open(fn, 'rt') as f:
for line in f:
parts = line.split('\t')
base_accession = parts[0]
if base_accession == 'accession':
continue
try:
taxid = parts[2]
except:
print(line)
taxid = int(taxid)
accession_lookup[base_accession] = taxid
log.info('Loaded accession to taxid mapping: %.2ss',
time.time() - start_time)
for i, header in enumerate(split_file(args.infile, '\x00')):
accessions = [
x.split(' ')[0].split('.')[0] for x in header.split('\x01')
]
taxids = []
for a in accessions:
try:
taxid = accession_lookup[a]
except KeyError:
log.info('Taxid for accession not found: %s', a)
taxids.append(taxid)
lca = tax_db.coverage_lca(taxids)
if not lca:
log.error('LCA not found for taxids %s', taxids)
lca = 1
print("{}\t{}".format(i + 1, lca), file=args.output)
def main():
parser = argparse.ArgumentParser(description='Get all viral taxons')
parser.add_argument('taxid_file')
parser.add_argument('--tax-dir', required=False)
args = parser.parse_args()
db = ncbitax.TaxonomyDb(tax_dir=args.tax_dir,
load_nodes=True,
load_names=True,
load_merged=True)
with open(args.taxid_file) as f:
for line in f:
taxid = int(line)
if db.is_viral(taxid):
print(taxid)
def prepare_karp_fasta(args):
tax_db = ncbitax.TaxonomyDb(tax_dir=args.tax_db,
load_names=True,
load_nodes=True)
if args.genomic_joined:
accession_lookup = {}
with open(args.genomic_joined, 'rt') as f:
for line in f:
parts = line.split('\t')
accession, taxid, gi = parts
taxid = int(taxid)
gi = int(gi)
accession_lookup[accession] = taxid
for seq in SeqIO.parse(args.infile, 'fasta'):
seq_id = seq.id
seq_id = seq_id.split(':')[0]
taxid = accession_lookup.get(seq_id)
if not taxid:
log.info('Sequence ID %s not found in taxonomy db', seq_id)
continue
karp_name = full_name(tax_db.parents, tax_db.names, taxid)
if not karp_name:
continue
seq.description = '\t'.join([seq.id, karp_name])
SeqIO.write(seq, args.output, 'fasta')
return
for seq in SeqIO.parse(args.infile, 'fasta'):
taxid = parse_taxid(seq.description)
if taxid == 0:
continue
# seq.id = str(taxid)
# seq.description = '\t'.join([str(taxid), full_name(tax_db.parents, tax_db.names, taxid)])
# seq.description = '\t'.join([str(taxid), seq.description])
karp_name = full_name(tax_db.parents, tax_db.names, taxid)
if not karp_name:
continue
seq.description = '\t'.join([seq.id, karp_name])
SeqIO.write(seq, args.output, 'fasta')
def compile_reports(args):
assert args.files or args.dir, "One of --files or --dir must be specified"
conf = load_config(args.config)
db = ncbitax.TaxonomyDb(tax_dir=args.tax_dir,
load_nodes=True,
load_names=True,
load_merged=True,
scientific_names_only=False)
with open(args.output, 'wt') as out_f:
writer = ReportTableWriter(out_f)
proc = ReportProcessor(db, conf)
proc.total_reads = {}
if args.read_counts:
with open(args.read_counts) as reads_f:
for row in csv.reader(reads_f, delimiter='\t'):
sample, count = row
if sample.endswith('.art'):
sample = sample[:-4]
proc.total_reads[sample] = int(count)
if args.classified_counts:
with open(args.classified_counts) as cc_f:
proc.classified_counts = read_classified_counts(cc_f)
taxa = []
if args.files:
for fn in args.files:
basename = os.path.basename(fn)
res = proc.process_file(basename, open(fn))
if res:
taxa.extend(res)
#sys.exit()
elif args.dir:
for x in Path(args.dir).iterdir():
if x.is_dir():
continue
basename = x.name
with x.open() as in_f:
res = proc.process_file(basename, in_f)
if res:
taxa.extend(res)
proc.sum_rank_abundances(taxa)
for d in taxa:
writer.write(d)
with open(args.rank_abundances, 'wt') as out_f:
for k, v in proc.rank_abundances.items():
parts = [str(x) for x in list(k) + [v]]
print('\t'.join(parts), file=out_f)
if args.missing_parents:
with open(args.missing_parents, 'wt') as f:
missing_parents = {
k
for k, v in proc.parent_path_cache.items() if v is None
}
log.info('Missing parents for %s taxids', len(missing_parents))
for taxid in sorted(list(missing_parents)):
if taxid == 0:
continue
print(taxid, file=f)
def create_metaothello_taxinfo(args):
taxids = set()
with open(args.taxids, 'rt') as f:
for line in f:
taxid = int(line)
# accession, taxid, gi = line.split('\t')
# taxid = int(taxid)
# accession, taxid, gi = line.split('\t')
# taxid = int(taxid)
# gi = int(gi)
taxids.add(taxid)
tax_db = ncbitax.TaxonomyDb(tax_dir=args.tax_dir,
load_nodes=True,
load_names=True,
load_merged=True)
taxid_info = {}
for taxid in taxids:
inf = taxid_info[taxid] = {
'species_taxid': taxid,
'species_name': tax_db.names[taxid]
}
while True:
taxid = tax_db.parents[taxid]
if taxid == 1:
break
if tax_db.ranks[taxid] == 'genus':
inf['genus_taxid'] = taxid
inf['genus_name'] = tax_db.names[taxid]
continue
elif tax_db.ranks[taxid] == 'family':
inf['family_taxid'] = taxid
inf['family_name'] = tax_db.names[taxid]
continue
elif tax_db.ranks[taxid] == 'order':
inf['order_taxid'] = taxid
inf['order_name'] = tax_db.names[taxid]
continue
elif tax_db.ranks[taxid] == 'class':
inf['class_taxid'] = taxid
inf['class_name'] = tax_db.names[taxid]
continue
elif tax_db.ranks[taxid] == 'phylum':
inf['phylum_taxid'] = taxid
inf['phylum_name'] = tax_db.names[taxid]
continue
elif tax_db.ranks[taxid] == 'kingdom':
inf['kingdom_taxid'] = taxid
inf['kingdom_name'] = tax_db.names[taxid]
continue
def create_index(key):
return {
taxid: i
for i, taxid in enumerate(
inf.get(key + '_taxid', -1) for inf in taxid_info.values())
}
species_index = create_index('species')
genus_index = create_index('genus')
family_index = create_index('family')
order_index = create_index('order')
class_index = create_index('class')
phylum_index = create_index('phylum')
kingdom_index = create_index('kingdom')
with open(args.output, 'wt') as f:
header = '\t'.join([
'Species_index', 'Species_ID', 'Species_name', 'Genus_index',
'Genus_ID', 'Genus_name', 'Family_index', 'Family_ID',
'Family_name', 'Order_index', 'Order_ID', 'Order_name',
'Class_index', 'Class_ID', 'Class_name', 'Phylum_index',
'Phylum_ID', 'Phylum_name'
])
print(header, file=f)
for taxid in taxids:
inf = taxid_info[taxid]
sid = inf.get('species_taxid', -1)
gid = inf.get('genus_taxid', -1)
fid = inf.get('family_taxid', -1)
oid = inf.get('order_taxid', -1)
cid = inf.get('class_taxid', -1)
pid = inf.get('phylum_taxid', -1)
parts = [
species_index[sid],
sid,
inf.get('species_name', 'NULL'),
genus_index[gid],
gid,
inf.get('genus_name', 'NULL'),
family_index[fid],
fid,
inf.get('family_name', 'NULL'),
order_index[oid],
oid,
inf.get('order_name', 'NULL'),
class_index[cid],
cid,
inf.get('class_name', 'NULL'),
phylum_index[pid],
pid,
inf.get('phylum_name', 'NULL'),
]
line = '\t'.join(str(x) for x in parts)
print(line, file=f)