def create_protein_alignment(protein_id, species): ''' Generates the SW alignment of three protein sequences: reference species protein, the assembled protein and the ensembl species protein @param protein_id: referent protein id @param species: species (latin) ''' sequences_for_fasta = [] dc = DirectoryCrawler() pc = ProteinContainer.Instance() dmc = DataMapContainer.Instance() acg = AlignmentCommandGenerator() tpc = TranslatedProteinContainer.Instance() data_map = dmc.get((protein_id, species)) # get all the proteins ref_protein = pc.get(protein_id) species_protein = pc.get(data_map.protein_id) assembled_protein = tpc.get(protein_id, species) sequences_for_fasta.append(ref_protein.get_sequence_record()) sequences_for_fasta.append(assembled_protein.get_sequence_record()) sequences_for_fasta.append(species_protein.get_sequence_record()) msa_fasta = "%s/%s.fa" % (dc.get_mafft_path(protein_id), species) msa_afa = "%s/%s.afa" % (dc.get_mafft_path(protein_id), species) msa_fasta_file = open(msa_fasta, "w") SeqIO.write(sequences_for_fasta, msa_fasta_file, "fasta") msa_fasta_file.close() mafft_cmd = acg.generate_mafft_command(msa_fasta, msa_afa) os.system(mafft_cmd)
def create_msa_alignments (): dc = DirectoryCrawler() pc = ProteinContainer.Instance() dmc = DataMapContainer.Instance() acg = AlignmentCommandGenerator() fill_all_containers(False) for (prot_id, exon_num) in get_protein_list(): if not check_status_file(prot_id): continue ref_prot_rec = pc.get(prot_id).get_sequence_record() exoloc_proteins = [] ensembl_proteins = [] exoloc_proteins.append(ref_prot_rec) ensembl_proteins.append(ref_prot_rec) assembled_dir = dc.get_assembled_protein_path(prot_id) for fasta in sorted(os.listdir(assembled_dir)): if fasta == "Homo_sapiens.fa": continue abs_fasta = "%s/%s" % (assembled_dir, fasta) prot_rec = load_fasta_single_record(abs_fasta, IUPAC.protein) exoloc_proteins.append(prot_rec) species_list = get_species_list(prot_id, None) for species in species_list: if species == "Homo_sapiens": continue data_map = dmc.get((prot_id, species)) prot_rec = pc.get(data_map.protein_id).get_sequence_record() prot_rec.id = species ensembl_proteins.append(prot_rec) msa_exoloc_path = "%s/msa_exoloc.fa" % dc.get_mafft_path(prot_id) msa_ensembl_path = "%s/msa_ensembl.fa" % dc.get_mafft_path(prot_id) write_seq_records_to_file(msa_exoloc_path, exoloc_proteins) write_seq_records_to_file(msa_ensembl_path, ensembl_proteins) cmd = acg.generate_mafft_command(msa_exoloc_path, "%s/msa_exoloc.afa" % dc.get_mafft_path(prot_id)) print cmd os.system(cmd) cmd = acg.generate_mafft_command(msa_ensembl_path, "%s/msa_ensembl.afa" % dc.get_mafft_path(prot_id)) print cmd os.system(cmd)
def create_species_msa_alignments (): dc = DirectoryCrawler() pc = ProteinContainer.Instance() dmc = DataMapContainer.Instance() acg = AlignmentCommandGenerator() fill_all_containers(False) for (prot_id, exon_num) in get_protein_list(): if not check_status_file(prot_id): continue ref_prot_rec = pc.get(prot_id).get_sequence_record() ref_prot_rec.id = "Homo_sapiens" assembled_dir = dc.get_assembled_protein_path(prot_id) species_list = get_species_list(prot_id, None) for species in species_list: protein_recs = [] protein_recs.append(ref_prot_rec) if species == "Homo_sapiens": continue data_map = dmc.get((prot_id, species)) prot_rec = pc.get(data_map.protein_id).get_sequence_record() prot_rec.id = species protein_recs.append(prot_rec) if "%s.fa" % species in os.listdir(assembled_dir): exoloc_protein_rec = load_fasta_single_record("%s/%s.fa" % (assembled_dir, species), IUPAC.protein) protein_recs.append(exoloc_protein_rec) msa_species_path = "%s/%s.fa" % (dc.get_mafft_path(prot_id), species) if len(protein_recs) == 1: continue write_seq_records_to_file(msa_species_path, protein_recs) cmd = acg.generate_mafft_command(msa_species_path, "%s/%s.afa" % (dc.get_mafft_path(prot_id), species)) print cmd os.system(cmd) os.remove(msa_species_path)
def main(): ''' Retrieves the list of all the proteins from reference species. For each ref species protein, it tries to find orthologues for all the species (from the species list) and generates the description file accordingly. If the description file already exists, it checks the status (OK/PARTIAL/FAILED). ''' reference_species = "Homo_sapiens" dc = DirectoryCrawler() acg = AlignmentCommandGenerator() logger = Logger.Instance() mutual_best_logger = logger.get_logger('mutual_best') protein_list = get_protein_list() species_list = get_default_species_list() failed_proteins = [] for (protein_id, num_of_exons) in protein_list: known_dict = {} abinitio_dict = {} print protein_id # generate all the directories for the protein dc.generate_directory_tree(protein_id) descr_file_path = dc.get_protein_description_file_path(protein_id) status_file_path = dc.get_mutual_best_status_file_path(protein_id) if (os.path.isfile(status_file_path) and os.path.getsize(status_file_path)): print DescriptionParser().get_protein_ids(protein_id) status_dict = read_status_file(protein_id) if (status_dict.has_key('MUTUAL_BEST')): if status_dict['MUTUAL_BEST'] == 'OK': mutual_best_logger.info('-,%s,mutual_best already exists for this protein - moving to the next one' % protein_id) else : mutual_best_logger.error('-,%s,mutual_best has failed for this protein (no orthologs found) - moving on the next one' % protein_id) failed_proteins.append(protein_id) continue # create the description file descr_file = open(descr_file_path, 'w') # reference protein file ref_species_pep = dc.get_protein_path(protein_id) + "/" + reference_species + ".fasta" fastacmd = acg.generate_fastacmd_protein_command(protein_id, reference_species, "all", ref_species_pep) p = Popen(fastacmd, shell=True, stdin=PIPE, stdout=PIPE, stderr=STDOUT, close_fds=True) output = p.stdout.read() if output: mutual_best_logger.error("%s,fastacmd error" % protein_id) # find orthologues for all species for species in species_list: find_ortholog_by_RBH (reference_species, species, ref_species_pep, protein_id, descr_file, mutual_best_logger) descr_file.close() mutual_best_logger.info("\n\n") # check what we've found out, whether this protein has any orthologs (known_dict, abinitio_dict) = DescriptionParser().get_protein_ids(protein_id) if (not abinitio_dict and (not known_dict or (len(known_dict.keys()) == 1 and known_dict.keys()[0] == reference_species))): mutual_best_logger.info ("-,%s, mutual best failed for this protein." % protein_id) update_entry_in_status_file(protein_id, "MUTUAL_BEST", "FAILED") failed_proteins.append(protein_id) else: update_entry_in_status_file(protein_id, "MUTUAL_BEST", "OK") print "Failed proteins: " for failed_protein_id in failed_proteins: print failed_protein_id
def _search_for_ortholog_in_database(original_species, target_species, original_protein_fasta, original_protein_id, db_type, logger): ''' Takes the reference species protein and makes a BLASTp query on the target_species database. Provided that there has been at least one hit, the best BLAST hit is queried against the reference species protein database. If the best hit from the second query is the original protein, then the protein data is returned. The queries as referred to as the forward and the backward hit respectively. @param db_type: all / abinitio ''' acg = AlignmentCommandGenerator() if (db_type == "all"): protein_id_pattern = re.compile("lcl\|(.*)\spep:(.*)\s(.*):(.*):(.*):(.*):(.*):(.*)\sgene:(.*)\stranscript:(.*)\s.*\s.*") else: protein_id_pattern = re.compile("lcl\|(.*)\spep:(.*)\s(.*):(.*):(.*):(.*):(.*):(.*)\stranscript:(.*)\s.*") output_file = "tmp.xml" forward_blastp_cmd = acg.generate_blastp_command_for_species(target_species, original_protein_fasta, output_file, db_type) print forward_blastp_cmd execute_command_and_log(logger, forward_blastp_cmd, (target_species, original_protein_id)) result_handle = open(output_file) blast_records = NCBIXML.parse(result_handle) try: best_forward_hit = blast_records.next() except (ValueError): logger.error("%s,%s,XML file empty - no forward blast results" % (target_species, original_protein_id)) return None if (best_forward_hit.alignments): bfh_title = _get_best_alignment(original_protein_id, best_forward_hit) protein_match = re.match(protein_id_pattern, bfh_title) else: return None if (db_type == "all"): (protein_id, protein_type, location_type, assembly, location_id, seq_start, seq_end, strand, gene_id, transcript_id) = protein_match.groups() else: (protein_id, protein_type, location_type, assembly, location_id, seq_start, seq_end, strand, transcript_id) = protein_match.groups() fasta_input_file = "species.fasta" fastacmd = acg.generate_fastacmd_protein_command(protein_id, target_species, db_type, fasta_input_file) execute_command_and_log(logger, fastacmd, (target_species, original_protein_id)) backward_blastp_cmd = acg.generate_blastp_command_for_species(original_species, fasta_input_file, output_file, "all") print backward_blastp_cmd execute_command_and_log(logger, backward_blastp_cmd, (target_species, original_protein_id)) result_handle = open(output_file) blast_records = NCBIXML.parse(result_handle) try: best_backward_hit = blast_records.next() except (ValueError): logger.error("%s,%s,XML file empty - no backward blast results" % (target_species, original_protein_id)) return None if (best_backward_hit.alignments): bbh_title = _get_best_alignment(original_protein_id, best_backward_hit) protein_match_b = re.match(protein_id_pattern, bbh_title) protein_id_b = protein_match_b.groups()[0] else: return None os.remove(output_file) os.remove(fasta_input_file) if (original_protein_id == protein_id_b): if (db_type == "all"): return (protein_id, protein_type, location_type, assembly, location_id, seq_start, seq_end, strand, gene_id, transcript_id) else: return (protein_id, protein_type, location_type, assembly, location_id, seq_start, seq_end, strand, transcript_id) else: return None
else: break i += 1 pattern = re.compile("lcl\|(.*)\spep::*") for title in best_alignments: prot_match = re.match(pattern, title) if prot_match.groups()[0] == protein_id: return title return best_alignments[0] if __name__ == '__main__': protein_id = "ENSP00000311134" acg = AlignmentCommandGenerator() dc = DirectoryCrawler() dc.generate_directory_tree(protein_id) descr_file_path = dc.get_protein_description_file_path(protein_id) descr_file = open(descr_file_path, 'w') output_file_path = dc.get_protein_path(protein_id) + "/" + "Homo_sapiens.fasta" fastacmd = acg.generate_fastacmd_protein_command(protein_id, "Homo_sapiens", "all", output_file_path) os.system(fastacmd) for species in get_default_species_list(): find_ortholog_by_RBH("Homo_sapiens", species, output_file_path, protein_id) descr_file.close()