def tblastn(input_fasta, blastdb, min_ident, min_covs, evalue, out_dir, blast_results_file, logging, num_threads=1, min_covhsp=25, seq_id_file=None): blast_runner = BlastRunner(input_fasta, out_dir) blast_runner.run_tblastn(query_fasta_path=input_fasta, blast_task='megablast', db_path=blastdb, db_type='protein', min_cov=min_covs, min_ident=min_ident, evalue=evalue, blast_outfile=blast_results_file, num_threads=num_threads, seq_id_file=seq_id_file, logging=logging) if os.path.getsize(blast_results_file) == 0: os.remove(blast_results_file) return False blast_df = BlastReader(blast_results_file, logging).df blast_df = blast_df.loc[blast_df['pident'] >= min_ident] blast_df = blast_df.loc[blast_df['qcovs'] >= min_covs] blast_df = blast_df.loc[blast_df['qcovhsp'] >= min_covhsp] blast_df = blast_df.loc[blast_df['evalue'] <= evalue] blast_df = fixStart(blast_df) blast_df = blast_df.sort_values(['sseqid', 'sstart', 'send', 'bitscore'], ascending=[True, True, True, False]) blast_df = blast_df.reset_index(drop=True) blast_df.to_csv(blast_results_file, sep='\t', header=True, line_terminator='\n', index=False) return True
def mob_blast(input_fasta, ref_db, min_ident, min_cov, evalue, tmp_dir, blast_results_file, overlap=5, num_threads=1): num_threads = 1 blast_runner = BlastRunner(input_fasta, tmp_dir) blast_runner.makeblastdb(ref_db, 'nucl') blast_runner.run_tblastn(query_fasta_path=input_fasta, blast_task='megablast', db_path=ref_db, db_type='nucl', min_cov=min_cov, min_ident=min_ident, evalue=evalue, blast_outfile=blast_results_file, num_threads=num_threads) if os.path.getsize(blast_results_file) == 0: return dict() blast_df = BlastReader(blast_results_file).df blast_df = blast_df.loc[blast_df['pident'] >= min_ident] blast_df = blast_df.loc[blast_df['qcovs'] >= min_cov] blast_df = blast_df.loc[blast_df['qcovhsp'] >= 25] blast_df = fixStart(blast_df) blast_df = blast_df.sort_values(['sseqid', 'sstart', 'send', 'bitscore'], ascending=[True, True, True, False]) blast_df = blast_df.reset_index(drop=True) blast_df = filter_overlaping_records(blast_df, overlap, 'sseqid', 'sstart', 'send', 'bitscore') prev_size = 0 size = str(len(blast_df)) while size != prev_size: blast_df = filter_overlaping_records(blast_df, overlap, 'sseqid', 'sstart', 'send', 'bitscore') prev_size = size size = str(len(blast_df)) #print(blast_df) return blast_df