def main():
import argparse
parser = argparse.ArgumentParser()
parser.add_argument('sigs', nargs='+')
parser.add_argument('lca_db')
args = parser.parse_args()
minhashes = []
for filename in args.sigs:
ss = sourmash.load_one_signature(filename)
minhashes.append(ss.minhash)
# load the LCA database
dblist, ksize, scaled = lca_utils.load_databases([args.lca_db], None)
db = dblist[0]
# double check scaled requirements
some_mh = minhashes[0]
mh_scaled = some_mh.scaled
if scaled >= mh_scaled:
print(
'** warning: many minhashes will go unclassified because LCA database scaled is {}'
.format(scaled),
file=sys.stderr)
print('** warning: the minhash scaled is {}'.format(mh_scaled),
file=sys.stderr)
summarize_taxonomic_purity(minhashes,
db,
verbose=True,
filenames=args.sigs)
def test_databases():
filename1 = utils.get_test_data('lca/delmont-1.lca.json')
filename2 = utils.get_test_data('lca/delmont-2.lca.json')
dblist, ksize, scaled = lca_utils.load_databases([filename1, filename2])
print(dblist)
assert len(dblist) == 2
assert ksize == 31
assert scaled == 10000
def main(args):
p = argparse.ArgumentParser()
p.add_argument('db', nargs='+')
p.add_argument('--scaled', type=float)
p.add_argument('-q',
'--quiet',
action='store_true',
help='suppress non-error output')
p.add_argument('-d',
'--debug',
action='store_true',
help='output debugging output')
p.add_argument(
'--minimum-num',
type=int,
default=0,
help=
'Minimum number of different lineages a k-mer must be in to be counted'
)
p.add_argument(
'--minimum-hashes',
type=int,
default=5,
help='Minimum number of hashes lineages must share to be reported')
p.add_argument('--lowest-rank', default='phylum')
p.add_argument('--prefix', default=None, help='prefix for output files')
args = p.parse_args(args)
if not args.db:
error('Error! must specify at least one LCA database with --db')
sys.exit(-1)
set_quiet(args.quiet, args.debug)
if args.scaled:
args.scaled = int(args.scaled)
# load all the databases
print('loading databases:', args.db)
dblist, ksize, scaled = lca_utils.load_databases(args.db, args.scaled)
assert len(dblist) == 1
# count all the LCAs across these databases
counts, confused_hashvals = make_lca_counts(dblist,
lowest_rank=args.lowest_rank,
min_num=args.minimum_num,
min_hashes=args.minimum_hashes,
prefix=args.prefix)
with open('confused_hashvals.txt', 'wt') as fp:
fp.write("\n".join([str(i) for i in confused_hashvals]))
def main():
import argparse
parser = argparse.ArgumentParser()
parser.add_argument('filename')
parser.add_argument('lca_db')
args = parser.parse_args()
# load the minhashes calculated by search.characterize_catlas_regions
group_ident = pickle.load(open(args.filename + '.node_mh', 'rb'))
# load the LCA database
dblist, ksize, scaled = lca_utils.load_databases([args.lca_db], None)
db = dblist[0]
# double check scaled requirements
some_mh = next(iter(group_ident.values()))
mh_scaled = some_mh.scaled
if scaled > mh_scaled:
print('** warning: many minhashes will go unclassified because LCA database scaled is {}'.format(scaled), file=sys.stderr)
print('** warning: the minhash scaled is {}'.format(mh_scaled), file=sys.stderr)
summarize_taxonomic_purity(group_ident.values(), db, verbose=True)
def main(args):
# load all the databases
dblist, ksize, scaled = lca_utils.load_databases(args.db, args.ksize,
args.scaled)
# count all the shared kmers across these databases
counts, total_kmer_count = count_shared_kmers(dblist)
# write out
with open(args.csv, 'wt') as fp:
hashes_by_lineage = csv.writer(fp)
hashes_by_lineage.writerow(
['rank', 'lineage', 'num_shared_kmers', 'percent_shared_kmers'])
for lineage, shared_kmer_count in counts.items():
rank = lineage[-1].rank
percent_shared_kmers = float(shared_kmer_count) / total_kmer_count
hashes_by_lineage.writerow([
rank,
lca_utils.display_lineage(lineage),
str(shared_kmer_count),
str(round(percent_shared_kmers, 2))
])