def profile_by_query(filename, genome_id, selected_loci, database):
# TODO: collect new alleles from here
allele_ids = ",".join("'{}'".format(operations.make_seqid(rec.seq)) for rec in SeqIO.parse(filename, "fasta"))
locus_ids = ",".join("'{}'".format(x) for x in selected_loci)
query = "select locus_id, allele_id from sequence where allele_id in ({}) and locus_id in ({});".format(allele_ids, locus_ids)
profile = sql_query(query, database=database).drop_duplicates("allele_id") # ensure allele_id is mapped only once
profile = profile.drop_duplicates("locus_id").set_index("locus_id") # ensure locus_id exists only once
profile = profile.rename(columns={"allele_id": genome_id}).iloc[:, 0]
return profile
def get(self, id):
if len(id) != 32:
abort(404)
sql = "select * from upload where batch_id='{}';".format(id)
results = db.sql_query(sql, database=DB).to_dict(orient="records")
if len(results) == 0:
abort(404)
Thread(target=internals.profiling_api,
args=(id, "Salmonella_5k", 95),
daemon=True).start()
return {"message": "Profiling dataset {}".format(id)}, 200
def get(self, id):
sql = None
if len(id) == 32:
sql = "select seq_id, batch_id, filename from upload where batch_id='{}';".format(
id)
elif len(id) == 64:
sql = "select seq_id, batch_id, filename from upload where seq_id='{}';".format(
id)
else:
abort(404)
results = db.sql_query(sql, database=DB).to_dict(orient="records")
if len(results) != 0:
return results
else:
abort(404)
def profiling(output_dir, input_dir, database, threads, occr_level=None, selected_loci=None, logger=None,
aligcov_cut=0.5, identity=90):
load_database_config()
if not logger:
logger = logs.console_logger(__name__)
logger.info("Renaming contigs...")
query_dir = files.joinpath(output_dir, "query")
files.create_if_not_exist(query_dir)
namemap = rename(query_dir, input_dir)
with open(files.joinpath(output_dir, "namemap.json"), "w") as f:
f.write(json.dumps(namemap))
if os.path.isdir(database):
logger.info("Profiling loci...")
refseq_fna = files.joinpath(database, "panRefSeq.fa")
profile_loci(refseq_fna, query_dir, output_dir, aligcov_cut, identity, threads)
logger.info("Allocating alleles...")
profile_alleles(query_dir, database, output_dir, threads, occr_level)
else:
logger.info("Identifying loci and allocating alleles...")
# select loci by scheme
if selected_loci:
selected_loci = set(selected_loci)
else:
query = "select locus_id from scheme where occurence>={};".format(occr_level)
selected_loci = set(sql_query(query, database=database).iloc[:, 0])
temp_dir = os.path.join(query_dir, "temp")
files.create_if_not_exist(temp_dir)
collect = []
args = [(os.path.join(query_dir, filename), temp_dir) for filename in os.listdir(query_dir) if filename.endswith(".fa")]
with ProcessPoolExecutor(threads) as executor:
for filename in executor.map(identify_loci, args):
genome_id = files.fasta_filename(filename)
target_file = os.path.join(temp_dir, genome_id + ".locus.fna")
profile = profile_by_query(target_file, genome_id, selected_loci, database)
collect.append(profile)
result = pd.concat(collect, axis=1)
result.to_csv(files.joinpath(output_dir, "wgmlst.tsv"), sep="\t")
shutil.rmtree(query_dir)
def get(self):
sql = "select seq_id, filename from upload;"
results = db.sql_query(sql, database=DB).to_dict(orient="records")
return results
def get(self):
sql = "select id from dendrogram;"
results = db.sql_query(sql, database=DB).to_dict(orient="records")
return results
def get(self):
sql = "select id, occurrence, database from profile;"
results = db.sql_query(sql, database=DB).to_dict(orient="records")
return results