aboutsummaryrefslogtreecommitdiff
path: root/scripts/download_genbank_data/from_genbank_to_fasta_and_yaml.py
blob: 4bb49646c5d361a586f1fe12313dfe552a31cd65 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
#!/usr/bin/env python3

import argparse
parser = argparse.ArgumentParser()
parser.add_argument('--skip-request', action='store_true', help='skip metadata and sequence request', required=False)
parser.add_argument('--only-missing-id', action='store_true', help='download only missing id', required=False)
parser.add_argument('--dict-ontology', type=str, help='where is the ontology',
                    default='../dict_ontology_standardization/',required=False)
args = parser.parse_args()

from Bio import Entrez
Entrez.email = 'another_email@gmail.com'

import xml.etree.ElementTree as ET
import json
import os
import requests
import sys

from datetime import date
from dateutil.parser import parse

num_ids_for_request = 100

dir_metadata = 'metadata_from_nuccore'
dir_fasta_and_yaml = 'fasta_and_yaml'
dir_dict_ontology_standardization = args.dict_ontology

today_date = date.today().strftime("%Y.%m.%d")
path_ncbi_virus_accession = 'sequences.{}.acc'.format(today_date)

def is_integer(string_to_check):
    try:
        int(string_to_check)
        return True
    except ValueError:
        return False

def chunks(lst, n):
    for i in range(0, len(lst), n):
        yield lst[i:i + n]

if os.path.exists(dir_metadata):
    print("The directory '{}' already exists.".format(dir_metadata))

    if not args.skip_request:
        print("\tTo start the request, delete the directory '{}' or specify --skip-request.".format(dir_metadata))
        sys.exit(-1)


accession_already_downloaded_set = []

if os.path.exists(dir_fasta_and_yaml):
    print("The directory '{}' already exists.".format(dir_fasta_and_yaml))
    if not args.only_missing_id:
        print("To start the download, delete the directory '{}' or specify --only-missing-id.".format(dir_fasta_and_yaml))
        sys.exit(-1)

    accession_already_downloaded_set = set([x.split('.yaml')[0].split('.')[0] for x in os.listdir(dir_fasta_and_yaml) if x.endswith('.yaml')])
    print('There are {} accession already downloaded.'.format(len(accession_already_downloaded_set)))


if not os.path.exists(dir_metadata):
    # Take all the ids
    id_set = set()

    # Try to search several strings
    term_list = ['SARS-CoV-2', 'SARS-CoV2', 'SARS CoV2', 'SARSCoV2', 'txid2697049[Organism]']
    for term in term_list:
        tmp_list = Entrez.read(
            Entrez.esearch(db='nuccore', term=term, idtype='acc', retmax='10000')
        )['IdList']

        # Remove mRNAs, ncRNAs, Proteins, and predicted models (more information here: https://en.wikipedia.org/wiki/RefSeq)
        tmp_list = [x for x in tmp_list if x[:2] not in ['NM', 'NR', 'NP', 'XM', 'XR', 'XP', 'WP']]

        # Remove the version in the id
        tmp_list = [x.split('.')[0] for x in tmp_list]

        #tmp_list = tmp_list[0:2] # restricting to small run
        new_ids_set = set([x.split('.')[0] for x in tmp_list])
        new_ids = len(new_ids_set.difference(id_set))
        id_set.update(new_ids_set)

        print('Term:', term, '-->', new_ids, 'new IDs from', len(tmp_list), '---> Total unique IDs:', len(id_set))

    if not os.path.exists(path_ncbi_virus_accession):
        r = requests.get('https://www.ncbi.nlm.nih.gov/genomes/VirusVariation/vvsearch2/?q=*:*&fq=%7B!tag=SeqType_s%7DSeqType_s:(%22Nucleotide%22)&fq=VirusLineageId_ss:(2697049)&cmd=download&sort=SourceDB_s%20desc,CreateDate_dt%20desc,id%20asc&dlfmt=acc&fl=id')
        with open(path_ncbi_virus_accession, 'w') as fw:
            fw.write(r.text)

    with open(path_ncbi_virus_accession) as f:
        tmp_list = [line.strip('\n') for line in f]

    new_ids = len(set(tmp_list).difference(id_set))
    id_set.update(tmp_list)

    print('DB: NCBI Virus', today_date, '-->', new_ids, 'new IDs from', len(tmp_list), '---> Total unique IDs:', len(id_set))

    if len(accession_already_downloaded_set) > 0:
        id_set = id_set.difference(accession_already_downloaded_set)
        print('There are {} missing IDs to download.'.format(len(id_set)))

    os.makedirs(dir_metadata)
    for i, id_x_list in enumerate(chunks(list(id_set), num_ids_for_request)):
        path_metadata_xxx_xml = os.path.join(dir_metadata, 'metadata_{}.xml'.format(i))
        print('Requesting {} ids --> {}'.format(len(id_x_list), path_metadata_xxx_xml))

        with open(path_metadata_xxx_xml, 'w') as fw:
            fw.write(
                Entrez.efetch(db='nuccore', id=id_x_list, retmode='xml').read()
            )


term_to_uri_dict = {}

for path_dict_xxx_csv in [os.path.join(dir_dict_ontology_standardization, name_xxx_csv) for name_xxx_csv in os.listdir(dir_dict_ontology_standardization) if name_xxx_csv.endswith('.csv')]:
    print('Read {}'.format(path_dict_xxx_csv))

    with open(path_dict_xxx_csv) as f:
        for line in f:
            if len(line.split(',')) > 2:
                term, uri = line.strip('\n').split('",')
                term = term.strip('"')
            else:
                term, uri = line.strip('\n').split(',')

            if term in term_to_uri_dict:
                print('Warning: in the dictionaries there are more entries for the same term ({}).'.format(term))
                continue

            term_to_uri_dict[term] = uri

if not os.path.exists(dir_fasta_and_yaml):
    os.makedirs(dir_fasta_and_yaml)

min_len_to_count = 27500
num_seq_with_len_ge_X_bp = 0

missing_value_list = []
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:
                info_for_yaml_dict['submitter']['authors'] = ["{}".format(x.text) for x in GBSeq_references.iter('GBAuthor')]

                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


            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'],
                    ['sequence_assembly_method', '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.strip('(average)').strip("reads/nt").strip('(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 term_to_uri_dict:
                                    seq_tec = term_to_uri_dict[seq_tec]
                                else:
                                    missing_value_list.append('\t'.join([accession_version, 'sample_sequencing_technology', seq_tec]))

                                new_seq_tec_list.append(seq_tec)

                            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_list = GBQualifier_value_text.split('; ')

                        if GBQualifier_value_text_list[0] in term_to_uri_dict:
                            info_for_yaml_dict['host']['host_species'] = term_to_uri_dict[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'] = term_to_uri_dict['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 term_to_uri_dict:
                                info_for_yaml_dict['host']['host_health_status'] = term_to_uri_dict[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 > -1:
                                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 term_to_uri_dict:
                            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 term_to_uri_dict:
                            info_for_yaml_dict['sample']['specimen_source'] = [term_to_uri_dict[GBQualifier_value_text]]
                        else:
                            if GBQualifier_value_text.lower() in ['np/op', 'np/op swab', 'np/np swab', 'nasopharyngeal and oropharyngeal swab', 'nasopharyngeal/oropharyngeal swab', 'combined nasopharyngeal and oropharyngeal swab']:
                                info_for_yaml_dict['sample']['specimen_source'] = [term_to_uri_dict['nasopharyngeal swab'], term_to_uri_dict['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']:
                                info_for_yaml_dict['sample']['specimen_source'] = [term_to_uri_dict['nasopharyngeal swab'], term_to_uri_dict['throat swab']]
                            elif GBQualifier_value_text.lower() in ['nasal swab and throat swab']:
                                info_for_yaml_dict['sample']['specimen_source'] = [term_to_uri_dict['nasal swab'], term_to_uri_dict['throat swab']]
                            elif GBQualifier_value_text.lower() in ['nasal-swab and oro-pharyngeal swab']:
                                info_for_yaml_dict['sample']['specimen_source'] = [term_to_uri_dict['nasal swab'], term_to_uri_dict['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 += '-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_value_text == 'Hong Kong':
                            GBQualifier_value_text = 'China: Hong Kong'

                        if GBQualifier_value_text in term_to_uri_dict:
                            info_for_yaml_dict['sample']['collection_location'] = term_to_uri_dict[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]


            # Remove technology key if empty!
            if (info_for_yaml_dict['technology']=={}):
                del info_for_yaml_dict['technology']

            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)


            if(len(GBSeq_sequence.text) >= min_len_to_count):
                num_seq_with_len_ge_X_bp += 1
        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.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.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))

print('Num. new sequences with length >= {} bp: {}'.format(min_len_to_count, num_seq_with_len_ge_X_bp))