From 6fd44b33eec7e72da5578ddef88a3ad18576bc1f Mon Sep 17 00:00:00 2001 From: Pjotr Prins Date: Sat, 2 Jan 2021 12:06:08 +0000 Subject: transform-genbank-xml2yamlfa.py rewrite --- workflows/pull-data/genbank/README.md | 2 +- workflows/pull-data/genbank/genbank.py | 339 ++++++++++++++++++ .../genbank/transform-genbank-xml2yamlfa.py | 383 ++++----------------- workflows/pull-data/genbank/update-from-genbank.py | 1 + 4 files changed, 403 insertions(+), 322 deletions(-) create mode 100644 workflows/pull-data/genbank/genbank.py (limited to 'workflows/pull-data/genbank') diff --git a/workflows/pull-data/genbank/README.md b/workflows/pull-data/genbank/README.md index 5597234..b5bac84 100644 --- a/workflows/pull-data/genbank/README.md +++ b/workflows/pull-data/genbank/README.md @@ -8,7 +8,7 @@ # --- fetch XML python3 update-from-genbank.py --ids genbank_ids.txt --out ~/tmp/genbank # --- Transform to YAML and FASTA -transform-genbank-xml2yamlfa --dir ~/tmp/genbank id --outdir ~/tmp/pubseq +python3 transform-genbank-xml2yamlfa --out ~/tmp/pubseq file(s) ``` # TODO diff --git a/workflows/pull-data/genbank/genbank.py b/workflows/pull-data/genbank/genbank.py new file mode 100644 index 0000000..7ce3913 --- /dev/null +++ b/workflows/pull-data/genbank/genbank.py @@ -0,0 +1,339 @@ +# Genbank XML parser + +import os +import sys +import xml.etree.ElementTree as ET + +class GBError(Exception): + pass + +def get_metadata(id, gb): + return True,None + +def get_sequence(id, gbseq): + seq = None + count = 0 + for gbseq_sequence in gbseq.findall('./GBSeq_sequence'): + count += 1 + if count > 1: + raise GBError(f"Expected one sequence for {id}") + seq = gbseq_sequence.text.upper() + print(f"SEQ: size={len(seq)}",seq[0:30]) + if len(seq) < 20_000: + raise GBError(f"Sequence too short") + return seq + +min_len_to_count = 15000 +num_seq_with_len_ge_X_bp = 0 + +missing_value_list = [] +not_created_accession_dict = {} +accession_with_errors_list = [] +if None: + + tree = ET.parse(path_metadata_xxx_xml) + GBSet = tree.getroot() + + for GBSeq in GBSet: + accession_version = GBSeq.find('GBSeq_accession-version').text + + GBSeq_sequence = GBSeq.find('GBSeq_sequence') + if GBSeq_sequence is None: + print(accession_version, ' - sequence not found') + continue + + try: + # print(path_metadata_xxx_xml, accession_version) + + # A general default-empty yaml could be read from the definitive one + info_for_yaml_dict = { + 'id': 'placeholder', + 'host': {}, + 'sample': {}, + 'virus': {}, + 'technology': {}, + 'submitter': {} + } + + + info_for_yaml_dict['sample']['sample_id'] = accession_version + info_for_yaml_dict['sample']['source_database_accession'] = ["http://identifiers.org/insdc/"+accession_version+"#sequence"] #accession is turned into resolvable URL/URI now + + + # submitter info + GBSeq_references = GBSeq.find('GBSeq_references') + if GBSeq_references is not None: + author_list = ["{}".format(x.text) for x in GBSeq_references.iter('GBAuthor')] + if len(author_list) > 0: + info_for_yaml_dict['submitter']['authors'] = author_list + + GBReference = GBSeq_references.find('GBReference') + if GBReference is not None: + GBReference_journal = GBReference.find('GBReference_journal') + + if GBReference_journal is not None and GBReference_journal.text != 'Unpublished': + if 'Submitted' in GBReference_journal.text: + info_for_yaml_dict['submitter']['submitter_name'] = ["{}".format(GBReference_journal.text.split(') ')[1].split(',')[0].strip())] + info_for_yaml_dict['submitter']['submitter_address'] = ','.join(GBReference_journal.text.split(') ')[1].split(',')[1:]).strip() + else: + info_for_yaml_dict['submitter']['additional_submitter_information'] = GBReference_journal.text + + # This script download and prepare data and metadata for assemblies samples + info_for_yaml_dict['technology']['assembly_method'] = 'http://purl.obolibrary.org/obo/GENEPIO_0001628' + + GBSeq_comment = GBSeq.find('GBSeq_comment') + if GBSeq_comment is not None and 'Assembly-Data' in GBSeq_comment.text: + prefix_split_string = '##Genome-Assembly' if GBSeq_comment.text.startswith('##Genome-') else '##Assembly' + + GBSeq_comment_text = GBSeq_comment.text.split( + '{}-Data-START## ; '.format(prefix_split_string) + )[1].split(' ; {}-Data-END##'.format(prefix_split_string))[0] + + for info_to_check, field_in_yaml in zip( + ['Assembly Method', 'Coverage', 'Sequencing Technology'], + ['alignment_protocol', 'sequencing_coverage', 'sample_sequencing_technology'] + ): + if info_to_check in GBSeq_comment_text: + tech_info_to_parse = GBSeq_comment_text.split('{} :: '.format(info_to_check))[1].split(' ;')[0] + + if field_in_yaml == 'sequencing_coverage': + # A regular expression would be better! + try: + info_for_yaml_dict['technology'][field_in_yaml] = [ + float(tech_info_to_parse.replace('(average)', '').replace("reads/nt", ''). + replace('(average for 6 sequences)', '').replace(',', '.').strip(' xX>')) + ] + except ValueError: + print(accession_version, "Couldn't make sense of Coverage '%s'" % tech_info_to_parse) + pass + elif field_in_yaml == 'sample_sequencing_technology': + new_seq_tec_list = [] + for seq_tec in tech_info_to_parse.split(';'): + seq_tec = seq_tec.strip() + if seq_tec in field_to_term_to_uri_dict['ncbi_sequencing_technology']: + seq_tec = field_to_term_to_uri_dict['ncbi_sequencing_technology'][seq_tec] + new_seq_tec_list.append(seq_tec) + else: + missing_value_list.append('\t'.join([accession_version, 'sample_sequencing_technology', seq_tec])) + + if len(new_seq_tec_list) > 0: + info_for_yaml_dict['technology']['sample_sequencing_technology'] = [x for x in new_seq_tec_list] + else: + info_for_yaml_dict['technology'][field_in_yaml] = tech_info_to_parse + + + for GBFeature in GBSeq.iter('GBFeature'): + if GBFeature.find('GBFeature_key').text != 'source': + continue + + for GBQualifier in GBFeature.iter('GBQualifier'): + GBQualifier_value = GBQualifier.find('GBQualifier_value') + if GBQualifier_value is None: + continue + GBQualifier_value_text = GBQualifier_value.text + + GBQualifier_name_text = GBQualifier.find('GBQualifier_name').text + + if GBQualifier_name_text == 'host': + GBQualifier_value_text = GBQualifier_value_text.split(';')[0] # For case like Homo sapiens;sex:female + if GBQualifier_value_text in field_to_term_to_uri_dict['ncbi_host_species']: + # Cases like 'Felis catus; Domestic Shorthair' + info_for_yaml_dict['host']['host_species'] = field_to_term_to_uri_dict['ncbi_host_species'][GBQualifier_value_text] + else: + GBQualifier_value_text_list = GBQualifier_value_text.split('; ') + + if GBQualifier_value_text_list[0] in field_to_term_to_uri_dict['ncbi_host_species']: + info_for_yaml_dict['host']['host_species'] = field_to_term_to_uri_dict['ncbi_host_species'][GBQualifier_value_text_list[0]] + elif GBQualifier_value_text_list[0] and ('MT215193' in accession_version or 'MT270814' in accession_version): + # Information checked manually from NCBI Virus + info_for_yaml_dict['host']['host_species'] = field_to_term_to_uri_dict['ncbi_host_species']['Canis lupus familiaris'] + else: + missing_value_list.append('\t'.join([accession_version, 'host_species', GBQualifier_value_text_list[0]])) + + # Possible cases: + # - Homo sapiens --> ['Homo sapiens'] + # - Homo sapiens; female --> ['Homo sapiens', 'female'] + # - Homo sapiens; female 63 --> ['Homo sapiens', 'female 63'] + # - Homo sapiens; female; age 40 --> ['Homo sapiens', 'female', 'age 40'] + # - Homo sapiens; gender: F; age: 61 --> ['Homo sapiens', 'gender: F', 'age: 61'] + # - Homo sapiens; gender: M; age: 68 --> ['Homo sapiens', 'gender: M', 'age: 68'] + # - Homo sapiens; hospitalized patient --> ['Homo sapiens', 'hospitalized patient'] + # - Homo sapiens; male --> ['Homo sapiens', 'male'] + # - Homo sapiens; male; 63 --> ['Homo sapiens', 'male', '63'] + # - Homo sapiens; male; age 29 --> ['Homo sapiens', 'male', 'age 29'] + # - Homo sapiens; symptomatic --> ['Homo sapiens', 'symptomatic'] + if len(GBQualifier_value_text_list) > 1: + host_sex = '' + if 'female' in GBQualifier_value_text_list[1]: + host_sex = 'female' + elif 'male' in GBQualifier_value_text_list[1]: + host_sex = 'male' + elif 'gender' in GBQualifier_value_text_list[1]: + host_sex_one_lecter = GBQualifier_value_text_list[1].split(':')[-1].strip() + if host_sex_one_lecter in ['F', 'M']: + host_sex = 'female' if host_sex_one_lecter == 'F' else 'male' + + if host_sex in ['male', 'female']: + info_for_yaml_dict['host']['host_sex'] = "http://purl.obolibrary.org/obo/PATO_0000384" if host_sex == 'male' else "http://purl.obolibrary.org/obo/PATO_0000383" + elif GBQualifier_value_text_list[1] in field_to_term_to_uri_dict['ncbi_host_health_status']: + info_for_yaml_dict['host']['host_health_status'] = field_to_term_to_uri_dict['ncbi_host_health_status'][GBQualifier_value_text_list[1]] + else: + missing_value_list.append('\t'.join([accession_version, 'host_sex or host_health_status', GBQualifier_value_text_list[1]])) + + # Host age + host_age = -1 + if len(GBQualifier_value_text_list[1].split(' ')) > 1 and is_integer(GBQualifier_value_text_list[1].split(' ')[-1]): + host_age = int(GBQualifier_value_text_list[1].split(' ')[-1]) + elif len(GBQualifier_value_text_list) > 2 and is_integer(GBQualifier_value_text_list[2].split(' ')[-1]): + host_age = int(GBQualifier_value_text_list[2].split(' ')[-1]) + + if host_age >= 0 and host_age < 110: + info_for_yaml_dict['host']['host_age'] = host_age + info_for_yaml_dict['host']['host_age_unit'] = 'http://purl.obolibrary.org/obo/UO_0000036' + elif len(GBQualifier_value_text_list) > 2: + missing_value_list.append('\t'.join([accession_version, 'host_age', GBQualifier_value_text_list[2]])) + elif GBQualifier_name_text == 'collected_by': + if any([x in GBQualifier_value_text.lower() for x in ['institute', 'hospital', 'city', 'center']]): + info_for_yaml_dict['sample']['collecting_institution'] = GBQualifier_value_text + else: + info_for_yaml_dict['sample']['collector_name'] = GBQualifier_value_text + elif GBQualifier_name_text == 'isolation_source': + if GBQualifier_value_text.upper() in field_to_term_to_uri_dict['ncbi_speciesman_source']: + GBQualifier_value_text = GBQualifier_value_text.upper() # For example, in case of 'usa: wa' + + # Little cleaning + GBQualifier_value_text = GBQualifier_value_text.strip("/'") + + if GBQualifier_value_text in field_to_term_to_uri_dict['ncbi_speciesman_source']: + info_for_yaml_dict['sample']['specimen_source'] = [field_to_term_to_uri_dict['ncbi_speciesman_source'][GBQualifier_value_text]] + else: + if GBQualifier_value_text.lower() in ['np/op', 'np-op', 'np/op swab', 'np/np swab', 'nasopharyngeal and oropharyngeal swab', 'nasopharyngeal/oropharyngeal swab', 'combined nasopharyngeal and oropharyngeal swab', 'naso and/or oropharyngeal swab']: + info_for_yaml_dict['sample']['specimen_source'] = [field_to_term_to_uri_dict['ncbi_speciesman_source']['nasopharyngeal swab'], field_to_term_to_uri_dict['ncbi_speciesman_source']['oropharyngeal swab']] + elif GBQualifier_value_text.lower() in ['nasopharyngeal swab/throat swab', 'nasopharyngeal/throat swab', 'nasopharyngeal swab and throat swab', 'nasal swab and throat swab', 'nasopharyngeal aspirate/throat swab', 'Nasopharyngeal/Throat']: + info_for_yaml_dict['sample']['specimen_source'] = [field_to_term_to_uri_dict['ncbi_speciesman_source']['nasopharyngeal swab'], field_to_term_to_uri_dict['ncbi_speciesman_source']['throat swab']] + elif GBQualifier_value_text.lower() in ['nasopharyngeal aspirate & throat swab', 'nasopharyngeal aspirate and throat swab']: + info_for_yaml_dict['sample']['specimen_source'] = [field_to_term_to_uri_dict['ncbi_speciesman_source']['nasopharyngeal aspirate'], field_to_term_to_uri_dict['ncbi_speciesman_source']['throat swab']] + elif GBQualifier_value_text.lower() in ['nasal swab and throat swab']: + info_for_yaml_dict['sample']['specimen_source'] = [field_to_term_to_uri_dict['ncbi_speciesman_source']['nasal swab'], field_to_term_to_uri_dict['ncbi_speciesman_source']['throat swab']] + elif GBQualifier_value_text.lower() in ['nasal-swab and oro-pharyngeal swab']: + info_for_yaml_dict['sample']['specimen_source'] = [field_to_term_to_uri_dict['ncbi_speciesman_source']['nasal swab'], field_to_term_to_uri_dict['ncbi_speciesman_source']['oropharyngeal swab']] + else: + missing_value_list.append('\t'.join([accession_version, 'specimen_source', GBQualifier_value_text])) + elif GBQualifier_name_text == 'collection_date': + # TO_DO: which format we will use? + date_to_write = GBQualifier_value_text + + if len(GBQualifier_value_text.split('-')) == 1: + if int(GBQualifier_value_text) < 2020: + date_to_write = "{}-12-15".format(GBQualifier_value_text) + else: + date_to_write = "{}-01-15".format(GBQualifier_value_text) + + if 'additional_collection_information' in info_for_yaml_dict['sample']: + info_for_yaml_dict['sample']['additional_collection_information'] += "; The 'collection_date' is estimated (the original date was: {})".format(GBQualifier_value_text) + else: + info_for_yaml_dict['sample']['additional_collection_information'] = "The 'collection_date' is estimated (the original date was: {})".format(GBQualifier_value_text) + elif len(GBQualifier_value_text.split('-')) == 2: + date_to_write = parse(GBQualifier_value_text).strftime('%Y-%m') + '-15' + + if 'additional_collection_information' in info_for_yaml_dict['sample']: + info_for_yaml_dict['sample']['additional_collection_information'] += "; The 'collection_date' is estimated (the original date was: {})".format(GBQualifier_value_text) + else: + info_for_yaml_dict['sample']['additional_collection_information'] = "The 'collection_date' is estimated (the original date was: {})".format(GBQualifier_value_text) + elif len(GBQualifier_value_text.split('-')) == 3: + GBQualifier_value_text_list = GBQualifier_value_text.split('-') + + if GBQualifier_value_text_list[1].isalpha(): + date_to_write = parse(GBQualifier_value_text).strftime('%Y-%m-%d') + + info_for_yaml_dict['sample']['collection_date'] = date_to_write + elif GBQualifier_name_text in ['lat_lon', 'country']: + if GBQualifier_name_text == 'country' and ': ' in GBQualifier_value_text: + GBQualifier_value_text = GBQualifier_value_text.replace(': ', ':') + + if GBQualifier_value_text in field_to_term_to_uri_dict['ncbi_countries']: + info_for_yaml_dict['sample']['collection_location'] = field_to_term_to_uri_dict['ncbi_countries'][GBQualifier_value_text] + else: + missing_value_list.append('\t'.join([accession_version, GBQualifier_name_text, GBQualifier_value_text])) + elif GBQualifier_name_text == 'note': + if 'additional_collection_information' in info_for_yaml_dict['sample']: + info_for_yaml_dict['sample']['additional_collection_information'] += '; ' + GBQualifier_value_text + else: + info_for_yaml_dict['sample']['additional_collection_information'] = GBQualifier_value_text + elif GBQualifier_name_text == 'isolate': + info_for_yaml_dict['virus']['virus_strain'] = GBQualifier_value_text + elif GBQualifier_name_text == 'db_xref': + info_for_yaml_dict['virus']['virus_species'] = "http://purl.obolibrary.org/obo/NCBITaxon_"+GBQualifier_value_text.split('taxon:')[1] + + # Check if mandatory fields are missing + if 'sample_sequencing_technology' not in info_for_yaml_dict['technology']: + # print(accession_version, ' - technology not found') + if accession_version not in not_created_accession_dict: + not_created_accession_dict[accession_version] = [] + not_created_accession_dict[accession_version].append('sample_sequencing_technology not found') + + if 'collection_location' not in info_for_yaml_dict['sample']: + if accession_version not in not_created_accession_dict: + not_created_accession_dict[accession_version] = [] + not_created_accession_dict[accession_version].append('collection_location not found') + + if 'collection_date' not in info_for_yaml_dict['sample']: + if accession_version not in not_created_accession_dict: + not_created_accession_dict[accession_version] = [] + not_created_accession_dict[accession_version].append('collection_date not found') + else: + year, month, day = [int(x) for x in info_for_yaml_dict['sample']['collection_date'].split('-')] + + collection_date_in_yaml = datetime(year, month, day) + if collection_date_in_yaml < min_acceptable_collection_date: + if accession_version not in not_created_accession_dict: + not_created_accession_dict[accession_version] = [] + not_created_accession_dict[accession_version].append('collection_date too early') + + if 'authors' not in info_for_yaml_dict['submitter']: + if accession_version not in not_created_accession_dict: + not_created_accession_dict[accession_version] = [] + not_created_accession_dict[accession_version].append('authors not found') + + if 'host_species' not in info_for_yaml_dict['host']: + if accession_version not in not_created_accession_dict: + not_created_accession_dict[accession_version] = [] + not_created_accession_dict[accession_version].append('host_species not found') + + if len(GBSeq_sequence.text) < min_len_to_count: + if accession_version not in not_created_accession_dict: + not_created_accession_dict[accession_version] = [] + not_created_accession_dict[accession_version].append('sequence shorter than {} bp'.format(min_len_to_count)) + + if accession_version not in not_created_accession_dict: + num_seq_with_len_ge_X_bp += 1 + + # with open(os.path.join(dir_fasta_and_yaml, '{}.fasta'.format(accession_version)), 'w') as fw: + # fw.write('>{}\n{}'.format(accession_version, GBSeq_sequence.text.upper())) + + with open(os.path.join(dir_fasta_and_yaml, '{}.yaml'.format(accession_version)), 'w') as fw: + json.dump(info_for_yaml_dict, fw, indent=2) + except: + print("Unexpected error for the ID {}: {}".format(accession_version, sys.exc_info()[0])) + accession_with_errors_list.append(accession_version) + continue + +if len(missing_value_list) > 0: + path_missing_terms_tsv = 'missing_terms.genbank.tsv' + print('Written missing terms in {}'.format(path_missing_terms_tsv)) + with open(path_missing_terms_tsv, 'w') as fw: + fw.write('\n'.join(missing_value_list)) + +if len(accession_with_errors_list) > 0: + path_accession_with_errors_tsv = 'accession_with_errors.genbank.tsv' + print('Written the accession with errors in {}'.format(path_accession_with_errors_tsv)) + with open(path_accession_with_errors_tsv, 'w') as fw: + fw.write('\n'.join(accession_with_errors_list)) + +if len(not_created_accession_dict) > 0: + path_not_created_accession_tsv = 'not_created_accession.genbank.tsv' + print('Written not created accession in {}'.format(path_not_created_accession_tsv)) + with open(path_not_created_accession_tsv, 'w') as fw: + fw.write('\n'.join(['\t'.join([accession_version, ','.join(missing_info_list)]) for accession_version, missing_info_list in not_created_accession_dict.items()])) + +print('Num. new sequences with length >= {} bp: {}'.format(min_len_to_count, num_seq_with_len_ge_X_bp)) diff --git a/workflows/pull-data/genbank/transform-genbank-xml2yamlfa.py b/workflows/pull-data/genbank/transform-genbank-xml2yamlfa.py index 8b970b6..c4e3eba 100755 --- a/workflows/pull-data/genbank/transform-genbank-xml2yamlfa.py +++ b/workflows/pull-data/genbank/transform-genbank-xml2yamlfa.py @@ -1,326 +1,67 @@ #!/usr/bin/env python3 # # Create a single YAML/FASTA from genbank XML - +# +# transform-genbank-xml2yamlfa --out ~/tmp/pubseq file(s) +# +# Also writes a validation file in the outdir named state.json +# +# Where --in can be a file or a directory # ---------------------------------------------------------------------- -""" -Generate metadata (YAML) and FASTA files for each accession -""" - -if not os.path.exists(dir_fasta_and_yaml): - os.makedirs(dir_fasta_and_yaml) - -min_len_to_count = 15000 -num_seq_with_len_ge_X_bp = 0 - -missing_value_list = [] -not_created_accession_dict = {} -accession_with_errors_list = [] - -for path_metadata_xxx_xml in [os.path.join(dir_metadata, name_metadata_xxx_xml) for name_metadata_xxx_xml in os.listdir(dir_metadata) if name_metadata_xxx_xml.endswith('.xml')]: - tree = ET.parse(path_metadata_xxx_xml) - GBSet = tree.getroot() - - for GBSeq in GBSet: - accession_version = GBSeq.find('GBSeq_accession-version').text - - GBSeq_sequence = GBSeq.find('GBSeq_sequence') - if GBSeq_sequence is None: - print(accession_version, ' - sequence not found') - continue - - try: - # print(path_metadata_xxx_xml, accession_version) - - # A general default-empty yaml could be read from the definitive one - info_for_yaml_dict = { - 'id': 'placeholder', - 'host': {}, - 'sample': {}, - 'virus': {}, - 'technology': {}, - 'submitter': {} - } - - - info_for_yaml_dict['sample']['sample_id'] = accession_version - info_for_yaml_dict['sample']['source_database_accession'] = ["http://identifiers.org/insdc/"+accession_version+"#sequence"] #accession is turned into resolvable URL/URI now - - - # submitter info - GBSeq_references = GBSeq.find('GBSeq_references') - if GBSeq_references is not None: - author_list = ["{}".format(x.text) for x in GBSeq_references.iter('GBAuthor')] - if len(author_list) > 0: - info_for_yaml_dict['submitter']['authors'] = author_list - - GBReference = GBSeq_references.find('GBReference') - if GBReference is not None: - GBReference_journal = GBReference.find('GBReference_journal') - - if GBReference_journal is not None and GBReference_journal.text != 'Unpublished': - if 'Submitted' in GBReference_journal.text: - info_for_yaml_dict['submitter']['submitter_name'] = ["{}".format(GBReference_journal.text.split(') ')[1].split(',')[0].strip())] - info_for_yaml_dict['submitter']['submitter_address'] = ','.join(GBReference_journal.text.split(') ')[1].split(',')[1:]).strip() - else: - info_for_yaml_dict['submitter']['additional_submitter_information'] = GBReference_journal.text - - # This script download and prepare data and metadata for assemblies samples - info_for_yaml_dict['technology']['assembly_method'] = 'http://purl.obolibrary.org/obo/GENEPIO_0001628' - - GBSeq_comment = GBSeq.find('GBSeq_comment') - if GBSeq_comment is not None and 'Assembly-Data' in GBSeq_comment.text: - prefix_split_string = '##Genome-Assembly' if GBSeq_comment.text.startswith('##Genome-') else '##Assembly' - - GBSeq_comment_text = GBSeq_comment.text.split( - '{}-Data-START## ; '.format(prefix_split_string) - )[1].split(' ; {}-Data-END##'.format(prefix_split_string))[0] - - for info_to_check, field_in_yaml in zip( - ['Assembly Method', 'Coverage', 'Sequencing Technology'], - ['alignment_protocol', 'sequencing_coverage', 'sample_sequencing_technology'] - ): - if info_to_check in GBSeq_comment_text: - tech_info_to_parse = GBSeq_comment_text.split('{} :: '.format(info_to_check))[1].split(' ;')[0] - - if field_in_yaml == 'sequencing_coverage': - # A regular expression would be better! - try: - info_for_yaml_dict['technology'][field_in_yaml] = [ - float(tech_info_to_parse.replace('(average)', '').replace("reads/nt", ''). - replace('(average for 6 sequences)', '').replace(',', '.').strip(' xX>')) - ] - except ValueError: - print(accession_version, "Couldn't make sense of Coverage '%s'" % tech_info_to_parse) - pass - elif field_in_yaml == 'sample_sequencing_technology': - new_seq_tec_list = [] - for seq_tec in tech_info_to_parse.split(';'): - seq_tec = seq_tec.strip() - if seq_tec in field_to_term_to_uri_dict['ncbi_sequencing_technology']: - seq_tec = field_to_term_to_uri_dict['ncbi_sequencing_technology'][seq_tec] - new_seq_tec_list.append(seq_tec) - else: - missing_value_list.append('\t'.join([accession_version, 'sample_sequencing_technology', seq_tec])) - - if len(new_seq_tec_list) > 0: - info_for_yaml_dict['technology']['sample_sequencing_technology'] = [x for x in new_seq_tec_list] - else: - info_for_yaml_dict['technology'][field_in_yaml] = tech_info_to_parse - - - for GBFeature in GBSeq.iter('GBFeature'): - if GBFeature.find('GBFeature_key').text != 'source': - continue - - for GBQualifier in GBFeature.iter('GBQualifier'): - GBQualifier_value = GBQualifier.find('GBQualifier_value') - if GBQualifier_value is None: - continue - GBQualifier_value_text = GBQualifier_value.text - - GBQualifier_name_text = GBQualifier.find('GBQualifier_name').text - - if GBQualifier_name_text == 'host': - GBQualifier_value_text = GBQualifier_value_text.split(';')[0] # For case like Homo sapiens;sex:female - if GBQualifier_value_text in field_to_term_to_uri_dict['ncbi_host_species']: - # Cases like 'Felis catus; Domestic Shorthair' - info_for_yaml_dict['host']['host_species'] = field_to_term_to_uri_dict['ncbi_host_species'][GBQualifier_value_text] - else: - GBQualifier_value_text_list = GBQualifier_value_text.split('; ') - - if GBQualifier_value_text_list[0] in field_to_term_to_uri_dict['ncbi_host_species']: - info_for_yaml_dict['host']['host_species'] = field_to_term_to_uri_dict['ncbi_host_species'][GBQualifier_value_text_list[0]] - elif GBQualifier_value_text_list[0] and ('MT215193' in accession_version or 'MT270814' in accession_version): - # Information checked manually from NCBI Virus - info_for_yaml_dict['host']['host_species'] = field_to_term_to_uri_dict['ncbi_host_species']['Canis lupus familiaris'] - else: - missing_value_list.append('\t'.join([accession_version, 'host_species', GBQualifier_value_text_list[0]])) - - # Possible cases: - # - Homo sapiens --> ['Homo sapiens'] - # - Homo sapiens; female --> ['Homo sapiens', 'female'] - # - Homo sapiens; female 63 --> ['Homo sapiens', 'female 63'] - # - Homo sapiens; female; age 40 --> ['Homo sapiens', 'female', 'age 40'] - # - Homo sapiens; gender: F; age: 61 --> ['Homo sapiens', 'gender: F', 'age: 61'] - # - Homo sapiens; gender: M; age: 68 --> ['Homo sapiens', 'gender: M', 'age: 68'] - # - Homo sapiens; hospitalized patient --> ['Homo sapiens', 'hospitalized patient'] - # - Homo sapiens; male --> ['Homo sapiens', 'male'] - # - Homo sapiens; male; 63 --> ['Homo sapiens', 'male', '63'] - # - Homo sapiens; male; age 29 --> ['Homo sapiens', 'male', 'age 29'] - # - Homo sapiens; symptomatic --> ['Homo sapiens', 'symptomatic'] - if len(GBQualifier_value_text_list) > 1: - host_sex = '' - if 'female' in GBQualifier_value_text_list[1]: - host_sex = 'female' - elif 'male' in GBQualifier_value_text_list[1]: - host_sex = 'male' - elif 'gender' in GBQualifier_value_text_list[1]: - host_sex_one_lecter = GBQualifier_value_text_list[1].split(':')[-1].strip() - if host_sex_one_lecter in ['F', 'M']: - host_sex = 'female' if host_sex_one_lecter == 'F' else 'male' - - if host_sex in ['male', 'female']: - info_for_yaml_dict['host']['host_sex'] = "http://purl.obolibrary.org/obo/PATO_0000384" if host_sex == 'male' else "http://purl.obolibrary.org/obo/PATO_0000383" - elif GBQualifier_value_text_list[1] in field_to_term_to_uri_dict['ncbi_host_health_status']: - info_for_yaml_dict['host']['host_health_status'] = field_to_term_to_uri_dict['ncbi_host_health_status'][GBQualifier_value_text_list[1]] - else: - missing_value_list.append('\t'.join([accession_version, 'host_sex or host_health_status', GBQualifier_value_text_list[1]])) - - # Host age - host_age = -1 - if len(GBQualifier_value_text_list[1].split(' ')) > 1 and is_integer(GBQualifier_value_text_list[1].split(' ')[-1]): - host_age = int(GBQualifier_value_text_list[1].split(' ')[-1]) - elif len(GBQualifier_value_text_list) > 2 and is_integer(GBQualifier_value_text_list[2].split(' ')[-1]): - host_age = int(GBQualifier_value_text_list[2].split(' ')[-1]) - - if host_age >= 0 and host_age < 110: - info_for_yaml_dict['host']['host_age'] = host_age - info_for_yaml_dict['host']['host_age_unit'] = 'http://purl.obolibrary.org/obo/UO_0000036' - elif len(GBQualifier_value_text_list) > 2: - missing_value_list.append('\t'.join([accession_version, 'host_age', GBQualifier_value_text_list[2]])) - elif GBQualifier_name_text == 'collected_by': - if any([x in GBQualifier_value_text.lower() for x in ['institute', 'hospital', 'city', 'center']]): - info_for_yaml_dict['sample']['collecting_institution'] = GBQualifier_value_text - else: - info_for_yaml_dict['sample']['collector_name'] = GBQualifier_value_text - elif GBQualifier_name_text == 'isolation_source': - if GBQualifier_value_text.upper() in field_to_term_to_uri_dict['ncbi_speciesman_source']: - GBQualifier_value_text = GBQualifier_value_text.upper() # For example, in case of 'usa: wa' - - # Little cleaning - GBQualifier_value_text = GBQualifier_value_text.strip("/'") - - if GBQualifier_value_text in field_to_term_to_uri_dict['ncbi_speciesman_source']: - info_for_yaml_dict['sample']['specimen_source'] = [field_to_term_to_uri_dict['ncbi_speciesman_source'][GBQualifier_value_text]] - else: - if GBQualifier_value_text.lower() in ['np/op', 'np-op', 'np/op swab', 'np/np swab', 'nasopharyngeal and oropharyngeal swab', 'nasopharyngeal/oropharyngeal swab', 'combined nasopharyngeal and oropharyngeal swab', 'naso and/or oropharyngeal swab']: - info_for_yaml_dict['sample']['specimen_source'] = [field_to_term_to_uri_dict['ncbi_speciesman_source']['nasopharyngeal swab'], field_to_term_to_uri_dict['ncbi_speciesman_source']['oropharyngeal swab']] - elif GBQualifier_value_text.lower() in ['nasopharyngeal swab/throat swab', 'nasopharyngeal/throat swab', 'nasopharyngeal swab and throat swab', 'nasal swab and throat swab', 'nasopharyngeal aspirate/throat swab', 'Nasopharyngeal/Throat']: - info_for_yaml_dict['sample']['specimen_source'] = [field_to_term_to_uri_dict['ncbi_speciesman_source']['nasopharyngeal swab'], field_to_term_to_uri_dict['ncbi_speciesman_source']['throat swab']] - elif GBQualifier_value_text.lower() in ['nasopharyngeal aspirate & throat swab', 'nasopharyngeal aspirate and throat swab']: - info_for_yaml_dict['sample']['specimen_source'] = [field_to_term_to_uri_dict['ncbi_speciesman_source']['nasopharyngeal aspirate'], field_to_term_to_uri_dict['ncbi_speciesman_source']['throat swab']] - elif GBQualifier_value_text.lower() in ['nasal swab and throat swab']: - info_for_yaml_dict['sample']['specimen_source'] = [field_to_term_to_uri_dict['ncbi_speciesman_source']['nasal swab'], field_to_term_to_uri_dict['ncbi_speciesman_source']['throat swab']] - elif GBQualifier_value_text.lower() in ['nasal-swab and oro-pharyngeal swab']: - info_for_yaml_dict['sample']['specimen_source'] = [field_to_term_to_uri_dict['ncbi_speciesman_source']['nasal swab'], field_to_term_to_uri_dict['ncbi_speciesman_source']['oropharyngeal swab']] - else: - missing_value_list.append('\t'.join([accession_version, 'specimen_source', GBQualifier_value_text])) - elif GBQualifier_name_text == 'collection_date': - # TO_DO: which format we will use? - date_to_write = GBQualifier_value_text - - if len(GBQualifier_value_text.split('-')) == 1: - if int(GBQualifier_value_text) < 2020: - date_to_write = "{}-12-15".format(GBQualifier_value_text) - else: - date_to_write = "{}-01-15".format(GBQualifier_value_text) - - if 'additional_collection_information' in info_for_yaml_dict['sample']: - info_for_yaml_dict['sample']['additional_collection_information'] += "; The 'collection_date' is estimated (the original date was: {})".format(GBQualifier_value_text) - else: - info_for_yaml_dict['sample']['additional_collection_information'] = "The 'collection_date' is estimated (the original date was: {})".format(GBQualifier_value_text) - elif len(GBQualifier_value_text.split('-')) == 2: - date_to_write = parse(GBQualifier_value_text).strftime('%Y-%m') + '-15' - - if 'additional_collection_information' in info_for_yaml_dict['sample']: - info_for_yaml_dict['sample']['additional_collection_information'] += "; The 'collection_date' is estimated (the original date was: {})".format(GBQualifier_value_text) - else: - info_for_yaml_dict['sample']['additional_collection_information'] = "The 'collection_date' is estimated (the original date was: {})".format(GBQualifier_value_text) - elif len(GBQualifier_value_text.split('-')) == 3: - GBQualifier_value_text_list = GBQualifier_value_text.split('-') - - if GBQualifier_value_text_list[1].isalpha(): - date_to_write = parse(GBQualifier_value_text).strftime('%Y-%m-%d') - - info_for_yaml_dict['sample']['collection_date'] = date_to_write - elif GBQualifier_name_text in ['lat_lon', 'country']: - if GBQualifier_name_text == 'country' and ': ' in GBQualifier_value_text: - GBQualifier_value_text = GBQualifier_value_text.replace(': ', ':') - - if GBQualifier_value_text in field_to_term_to_uri_dict['ncbi_countries']: - info_for_yaml_dict['sample']['collection_location'] = field_to_term_to_uri_dict['ncbi_countries'][GBQualifier_value_text] - else: - missing_value_list.append('\t'.join([accession_version, GBQualifier_name_text, GBQualifier_value_text])) - elif GBQualifier_name_text == 'note': - if 'additional_collection_information' in info_for_yaml_dict['sample']: - info_for_yaml_dict['sample']['additional_collection_information'] += '; ' + GBQualifier_value_text - else: - info_for_yaml_dict['sample']['additional_collection_information'] = GBQualifier_value_text - elif GBQualifier_name_text == 'isolate': - info_for_yaml_dict['virus']['virus_strain'] = GBQualifier_value_text - elif GBQualifier_name_text == 'db_xref': - info_for_yaml_dict['virus']['virus_species'] = "http://purl.obolibrary.org/obo/NCBITaxon_"+GBQualifier_value_text.split('taxon:')[1] - - # Check if mandatory fields are missing - if 'sample_sequencing_technology' not in info_for_yaml_dict['technology']: - # print(accession_version, ' - technology not found') - if accession_version not in not_created_accession_dict: - not_created_accession_dict[accession_version] = [] - not_created_accession_dict[accession_version].append('sample_sequencing_technology not found') - - if 'collection_location' not in info_for_yaml_dict['sample']: - if accession_version not in not_created_accession_dict: - not_created_accession_dict[accession_version] = [] - not_created_accession_dict[accession_version].append('collection_location not found') - - if 'collection_date' not in info_for_yaml_dict['sample']: - if accession_version not in not_created_accession_dict: - not_created_accession_dict[accession_version] = [] - not_created_accession_dict[accession_version].append('collection_date not found') - else: - year, month, day = [int(x) for x in info_for_yaml_dict['sample']['collection_date'].split('-')] - - collection_date_in_yaml = datetime(year, month, day) - if collection_date_in_yaml < min_acceptable_collection_date: - if accession_version not in not_created_accession_dict: - not_created_accession_dict[accession_version] = [] - not_created_accession_dict[accession_version].append('collection_date too early') - - if 'authors' not in info_for_yaml_dict['submitter']: - if accession_version not in not_created_accession_dict: - not_created_accession_dict[accession_version] = [] - not_created_accession_dict[accession_version].append('authors not found') - - if 'host_species' not in info_for_yaml_dict['host']: - if accession_version not in not_created_accession_dict: - not_created_accession_dict[accession_version] = [] - not_created_accession_dict[accession_version].append('host_species not found') - - if len(GBSeq_sequence.text) < min_len_to_count: - if accession_version not in not_created_accession_dict: - not_created_accession_dict[accession_version] = [] - not_created_accession_dict[accession_version].append('sequence shorter than {} bp'.format(min_len_to_count)) - - if accession_version not in not_created_accession_dict: - num_seq_with_len_ge_X_bp += 1 - - with open(os.path.join(dir_fasta_and_yaml, '{}.fasta'.format(accession_version)), 'w') as fw: - fw.write('>{}\n{}'.format(accession_version, GBSeq_sequence.text.upper())) - - with open(os.path.join(dir_fasta_and_yaml, '{}.yaml'.format(accession_version)), 'w') as fw: - json.dump(info_for_yaml_dict, fw, indent=2) - except: - print("Unexpected error for the ID {}: {}".format(accession_version, sys.exc_info()[0])) - accession_with_errors_list.append(accession_version) - continue - -if len(missing_value_list) > 0: - path_missing_terms_tsv = 'missing_terms.genbank.tsv' - print('Written missing terms in {}'.format(path_missing_terms_tsv)) - with open(path_missing_terms_tsv, 'w') as fw: - fw.write('\n'.join(missing_value_list)) - -if len(accession_with_errors_list) > 0: - path_accession_with_errors_tsv = 'accession_with_errors.genbank.tsv' - print('Written the accession with errors in {}'.format(path_accession_with_errors_tsv)) - with open(path_accession_with_errors_tsv, 'w') as fw: - fw.write('\n'.join(accession_with_errors_list)) - -if len(not_created_accession_dict) > 0: - path_not_created_accession_tsv = 'not_created_accession.genbank.tsv' - print('Written not created accession in {}'.format(path_not_created_accession_tsv)) - with open(path_not_created_accession_tsv, 'w') as fw: - fw.write('\n'.join(['\t'.join([accession_version, ','.join(missing_info_list)]) for accession_version, missing_info_list in not_created_accession_dict.items()])) -print('Num. new sequences with length >= {} bp: {}'.format(min_len_to_count, num_seq_with_len_ge_X_bp)) +# See also directory .guix-run and README.md + +import argparse +import gzip +import os +import sys +import types +import xml.etree.ElementTree as ET +from utils import chunks +import genbank + +parser = argparse.ArgumentParser() +parser.add_argument('--out', type=str, help='Directory to write to', +required=True) +parser.add_argument('files', nargs='+', help='file(s)') +args = parser.parse_args() + +dir = args.out +if not os.path.exists(dir): + raise Exception(f"Directory {dir} does not exist") + +states = {} + +for xmlfn in args.files: + print(f"--- Reading {xmlfn}") + with gzip.open(xmlfn, 'r') as f: + xml = f.read().decode() + tree = ET.fromstring(xml) + for gb in tree.findall('./GBSeq'): + valid = None + error = None + id = gb.find("GBSeq_locus").text + basename = dir+"/"+id + print(f" parsing {id}") + try: + valid,meta = genbank.get_metadata(id,gb) + if valid: + fa = basename+".fa" + seq = genbank.get_sequence(id,gb) + print(f" writing {fa}") + with open(fa,"w") as f2: + f2.write(f"> {id}\n") + f2.write(seq) + # print(seq) + except genbank.GBError as e: + print(f"OS error: {e}") + valid = False + error = str(e) + state = {} + if not valid: + state['valid'] = False + if error: + state['error'] = error + states[id] = state + +print(states) diff --git a/workflows/pull-data/genbank/update-from-genbank.py b/workflows/pull-data/genbank/update-from-genbank.py index 3faea39..dca5563 100755 --- a/workflows/pull-data/genbank/update-from-genbank.py +++ b/workflows/pull-data/genbank/update-from-genbank.py @@ -22,6 +22,7 @@ parser = argparse.ArgumentParser() parser.add_argument('--max', type=int, help='Max queries', required=False) parser.add_argument('--ids', type=str, help='File with ids to fetch, 1 id per line', required=True) parser.add_argument('--out', type=str, help='Directory to write to', required=True) + args = parser.parse_args() ids = set() -- cgit v1.2.3