def lineage_id(self, acc_number_list): """Get taxonomic lineage name for accession ids Given a list of accession numbers, yield the accession number and their associated lineage (in the form of taxids) as tuples Args: acc_number_list (:obj:`list`): a list of accession numbers Yields: tuple: (accession id, lineage list) """ self.check_list_ids(acc_number_list) with self.db.atomic(): query = Accession.select().where( Accession.accession << acc_number_list) for i in query: lineage_list = [] current_lineage = i.taxid.tax_name current_lineage_id = i.taxid.ncbi_taxid parent = i.taxid.parent_taxid while current_lineage != 'root': lineage_list.append(current_lineage_id) new_query = Taxa.get(Taxa.ncbi_taxid == parent) current_lineage = new_query.tax_name current_lineage_id = new_query.ncbi_taxid parent = new_query.parent_taxid yield (i.accession, lineage_list)
def sci_name(self, acc_number_list): """Get taxonomic scientific name for accession ids Given a list of accession numbers, yield the accession number and their associated scientific name as tuples Args: acc_number_list (:obj:`list`): a list of accession numbers Yields: tuple: (accession id, taxonomy id) """ self.check_list_ids(acc_number_list) with self.db.atomic(): query = Accession.select().where( Accession.accession << acc_number_list) for i in query: yield (i.accession, i.taxid.tax_name)
def accession2taxid(self, acc2taxid=None, chunk=None): """Parses the accession2taxid files This method parses the accession2taxid file, build a dictionary, stores it in a list and yield for insertion in the database. :: { 'accession': accession_id_from_file, 'taxid': associated_taxonomic_id } Args: acc2taxid (:obj:`str`): Path to acc2taxid input file (gzipped) chunk (:obj:`int`): Chunk size of entries to gather before yielding. Default 500 (set at object construction) Yields: list: Chunk size of read entries """ # Some accessions (e.g.: AAA22826) have a taxid = 0 entries = [] counter = 0 taxids = self.cache_taxids() if not self.fast: accessions = {} if acc2taxid is None: acc2taxid = self.acc_file self.check_file(acc2taxid) if chunk is None: chunk = self.chunk self.logger.debug("Parsing %s" % str(acc2taxid)) self.logger.debug("Fast mode %s" % "ON" if self.fast else "OFF") with gzip.open(acc2taxid, 'rb') as f: f.readline() # discard the header for line in f: line_list = line.decode().rstrip('\n').split('\t') # Check the taxid already exists and get its id if line_list[2] not in taxids: continue # In case of an update or parsing an already inserted list of # accessions if not self.fast: if line_list[0] in accessions: continue try: Accession.get(Accession.accession == line_list[0]) except Accession.DoesNotExist: accessions[line_list[0]] = True data_dict = { 'accession': line_list[0], 'taxid': line_list[2] } else: data_dict = { 'accession': line_list[0], 'taxid': line_list[2] } entries.append(data_dict) counter += 1 if counter == chunk: yield(entries) entries = [] counter = 0 if len(entries): yield(entries)
def accession2taxid(self, acc2taxid=None, chunk=None): """Parses the accession2taxid files This method parses the accession2taxid file, build a dictionary, stores it in a list and yield for insertion in the database. :: { 'accession': accession_id_from_file, 'taxid': associated_taxonomic_id } Args: acc2taxid (:obj:`str`): Path to acc2taxid input file (gzipped) chunk (:obj:`int`): Chunk size of entries to gather before yielding. Default 500 (set at object construction) Yields: list: Chunk size of read entries """ # Some accessions (e.g.: AAA22826) have a taxid = 0 entries = [] counter = 0 taxids = self.cache_taxids() if not self.fast: accessions = {} if acc2taxid is None: acc2taxid = self.acc_file self.check_file(acc2taxid) if chunk is None: chunk = self.chunk self.verbose("Parsing %s" % str(acc2taxid)) self.verbose("Fast mode %s" % "ON" if self.fast else "OFF") with gzip.open(acc2taxid, 'rb') as f: f.readline() # discard the header for line in f: line_list = line.decode().rstrip('\n').split('\t') # Check the taxid already exists and get its id if line_list[2] not in taxids: continue # In case of an update or parsing an already inserted list of # accessions if not self.fast: if line_list[0] in accessions: continue try: Accession.get(Accession.accession == line_list[0]) except Accession.DoesNotExist: accessions[line_list[0]] = True data_dict = { 'accession': line_list[0], 'taxid': line_list[2] } else: data_dict = { 'accession': line_list[0], 'taxid': line_list[2] } entries.append(data_dict) counter += 1 if counter == chunk: yield (entries) entries = [] counter = 0 if len(entries): yield (entries)
def create_db(args): """Main function for the 'taxadb create' sub-command. This function creates a taxonomy database with 2 tables: Taxa and Sequence. Args: args.input (:obj:`str`): input directory. It is the directory created by `taxadb download` args.dbname (:obj:`str`): name of the database to be created args.dbtype (:obj:`str`): type of database to be used. args.division (:obj:`str`): division to create the db for. args.fast (:obj:`bool`): Disables checks for faster db creation. Use with caution! """ database = DatabaseFactory(**args.__dict__).get_database() div = args.division # am lazy at typing db.initialize(database) nucl_est = 'nucl_est.accession2taxid.gz' nucl_gb = 'nucl_gb.accession2taxid.gz' nucl_gss = 'nucl_gss.accession2taxid.gz' nucl_wgs = 'nucl_wgs.accession2taxid.gz' prot = 'prot.accession2taxid.gz' acc_dl_list = [] db.connect() parser = TaxaDumpParser(nodes_file=os.path.join(args.input, 'nodes.dmp'), names_file=os.path.join(args.input, 'names.dmp'), verbose=args.verbose) parser.verbose("Connected to database ...") # If taxa table already exists, do not recreate and fill it # safe=True prevent not to create the table if it already exists if not Taxa.table_exists(): parser.verbose("Creating table %s" % str(Taxa.get_table_name())) db.create_table(Taxa, safe=True) parser.verbose("Parsing files") taxa_info_list = parser.taxdump() parser.verbose("Inserting taxa data") with db.atomic(): for i in range(0, len(taxa_info_list), args.chunk): Taxa.insert_many(taxa_info_list[i:i + args.chunk]).execute() print('Taxa: completed') parser.verbose("Checking table accession ...") # At first load, table accession does not exist yet, we create it db.create_table(Accession, safe=True) if div in ['full', 'nucl', 'est']: acc_dl_list.append(nucl_est) if div in ['full', 'nucl', 'gb']: acc_dl_list.append(nucl_gb) if div in ['full', 'nucl', 'gss']: acc_dl_list.append(nucl_gss) if div in ['full', 'nucl', 'wgs']: acc_dl_list.append(nucl_wgs) if div in ['full', 'prot']: acc_dl_list.append(prot) parser = Accession2TaxidParser(verbose=args.verbose, fast=args.fast) with db.atomic(): for acc_file in acc_dl_list: inserted_rows = 0 parser.verbose("Parsing %s" % str(acc_file)) for data_dict in parser.accession2taxid(acc2taxid=os.path.join( args.input, acc_file), chunk=args.chunk): Accession.insert_many(data_dict[0:args.chunk]).execute() inserted_rows += len(data_dict) print('%s: %s added to database (%d rows inserted)' % (Accession.get_table_name(), acc_file, inserted_rows)) if not Accession.has_index(name='accession_accession'): print('Creating index for %s' % Accession.get_table_name()) try: db.create_index(Accession, ['accession'], unique=True) except PeeweeException as err: raise Exception("Could not create Accession index: %s" % str(err)) print('Accession: completed') db.close()
def create_db(args): """Main function for the 'taxadb create' sub-command. This function creates a taxonomy database with 2 tables: Taxa and Sequence. Args: args.input (:obj:`str`): input directory. It is the directory created by `taxadb download` args.dbname (:obj:`str`): name of the database to be created args.dbtype (:obj:`str`): type of database to be used. args.division (:obj:`str`): division to create the db for. args.fast (:obj:`bool`): Disables checks for faster db creation. Use with caution! """ logger = logging.getLogger(__name__) database = DatabaseFactory(**args.__dict__).get_database() div = args.division # am lazy at typing db.initialize(database) nucl_gb = 'nucl_gb.accession2taxid.gz' nucl_wgs = 'nucl_wgs.accession2taxid.gz' prot = 'prot.accession2taxid.gz' acc_dl_list = [] db.connect() parser = TaxaDumpParser(nodes_file=os.path.join(args.input, 'nodes.dmp'), names_file=os.path.join(args.input, 'names.dmp'), verbose=args.verbose) logger.debug('Connected to database') # If taxa table already exists, do not recreate and fill it # safe=True prevent not to create the table if it already exists if not Taxa.table_exists(): logger.info('Creating table %s' % str(Taxa.get_table_name())) db.create_tables([Taxa]) logger.info("Parsing files") taxa_info_list = parser.taxdump() logger.info("Inserting taxonomy data") total_size = len(taxa_info_list) try: with db.atomic(): for i in tqdm(range(0, total_size, args.chunk), unit=' chunks', desc='INFO:taxadb.app', total=''): Taxa.insert_many(taxa_info_list[i:i + args.chunk]).execute() except OperationalError as e: print("\n") # needed because the above counter has none logger.error("sqlite3 error: %s" % e) logger.error("Maybe retry with a lower chunk size.") sys.exit(1) logger.info('Table Taxa completed') # At first load, table accession does not exist yet, we create it db.create_tables([Accession]) if div in ['full', 'nucl', 'gb']: acc_dl_list.append(nucl_gb) if div in ['full', 'nucl', 'wgs']: acc_dl_list.append(nucl_wgs) if div in ['full', 'prot']: acc_dl_list.append(prot) parser = Accession2TaxidParser(verbose=args.verbose, fast=args.fast) with db.atomic(): for acc_file in acc_dl_list: inserted_rows = 0 logger.info("Parsing %s" % str(acc_file)) for data_dict in tqdm(parser.accession2taxid( acc2taxid=os.path.join(args.input, acc_file), chunk=args.chunk), unit=' chunks', desc='INFO:taxadb.app', total=''): Accession.insert_many(data_dict[0:args.chunk]).execute() inserted_rows += len(data_dict) logger.info('%s: %s added to database (%d rows inserted)' % (Accession.get_table_name(), acc_file, inserted_rows)) if not Accession.has_index(name='accession_accession'): logger.info('Creating index for %s' % Accession.get_table_name()) try: # db.add_index(Accession, ['accession'], unique=True) idx = db.index(db.Accession, name='accession', unique=True) db.add_index(idx) except PeeweeException as err: raise Exception("Could not create Accession index: %s" % str(err)) logger.info('Table Accession completed') db.close()
def create_db(args): """Main function for the 'taxadb create' sub-command. This function creates a taxonomy database with 2 tables: Taxa and Sequence. Args: args.input (:obj:`str`): input directory. It is the directory created by `taxadb download` args.dbname (:obj:`str`): name of the database to be created args.dbtype (:obj:`str`): type of database to be used. args.division (:obj:`str`): division to create the db for. args.fast (:obj:`bool`): Disables checks for faster db creation. Use with caution! """ logger = logging.getLogger(__name__) database = DatabaseFactory(**args.__dict__).get_database() div = args.division # am lazy at typing db.initialize(database) nucl_gb = 'nucl_gb.accession2taxid.gz' nucl_wgs = 'nucl_wgs.accession2taxid.gz' prot = 'prot.accession2taxid.gz' acc_dl_list = [] db.connect() parser = TaxaDumpParser(nodes_file=os.path.join(args.input, 'nodes.dmp'), names_file=os.path.join(args.input, 'names.dmp'), verbose=args.verbose) logger.debug('Connected to database') # If taxa table already exists, do not recreate and fill it # safe=True prevent not to create the table if it already exists if not Taxa.table_exists(): logger.info('Creating table %s' % str(Taxa.get_table_name())) db.create_tables([Taxa]) logger.info("Parsing files") taxa_info_list = parser.taxdump() logger.info("Inserting taxonomy data") total_size = len(taxa_info_list) try: with db.atomic(): for i in tqdm(range(0, total_size, args.chunk), unit=' chunks', desc='INFO:taxadb.app', total=''): Taxa.insert_many(taxa_info_list[i:i+args.chunk]).execute() except OperationalError as e: print("\n") # needed because the above counter has none logger.error("sqlite3 error: %s" % e) logger.error("Maybe retry with a lower chunk size.") sys.exit(1) logger.info('Table Taxa completed') # At first load, table accession does not exist yet, we create it db.create_tables([Accession]) if div in ['full', 'nucl', 'gb']: acc_dl_list.append(nucl_gb) if div in ['full', 'nucl', 'wgs']: acc_dl_list.append(nucl_wgs) if div in ['full', 'prot']: acc_dl_list.append(prot) parser = Accession2TaxidParser(verbose=args.verbose, fast=args.fast) with db.atomic(): for acc_file in acc_dl_list: inserted_rows = 0 logger.info("Parsing %s" % str(acc_file)) for data_dict in tqdm( parser.accession2taxid( acc2taxid=os.path.join(args.input, acc_file), chunk=args.chunk), unit=' chunks', desc='INFO:taxadb.app', total=''): Accession.insert_many(data_dict[0:args.chunk]).execute() inserted_rows += len(data_dict) logger.info('%s: %s added to database (%d rows inserted)' % (Accession.get_table_name(), acc_file, inserted_rows)) if not Accession.has_index(name='accession_accession'): logger.info('Creating index for %s' % Accession.get_table_name()) try: # db.add_index(Accession, ['accession'], unique=True) idx = db.index(db.Accession, name='accession', unique=True) db.add_index(idx) except PeeweeException as err: raise Exception("Could not create Accession index: %s" % str(err)) logger.info('Table Accession completed') db.close()