-
Notifications
You must be signed in to change notification settings - Fork 4
/
upimapi.py
1434 lines (1263 loc) · 63.7 KB
/
upimapi.py
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
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
#!/usr/bin/env python
"""
UPIMAPI - UniProt Id Mapping through API
By João Sequeira
Mar 2020
"""
import json
from argparse import ArgumentParser, ArgumentTypeError
import os
import sys
from time import strftime, gmtime, time, sleep
from subprocess import run, Popen, PIPE, check_output
import requests
import yaml
from psutil import virtual_memory
from pathlib import Path
from multiprocessing import cpu_count, Pool, Manager
from io import StringIO
import pandas as pd
import xml.etree.ElementTree as ET
from tqdm import tqdm
from datetime import datetime
from Bio import SwissProt as SP
import numpy as np
from functools import partial
import re
__version__ = '1.13.1'
def load_api_info():
return yaml.safe_load(requests.get('https://rest.uniprot.org/docs/uniprot-openapi3.yaml').text)
def get_url(url, **kwargs):
response = requests.get(url, **kwargs)
if not response.ok:
response.raise_for_status()
sys.exit(response.txt)
return response
def get_uniprot_columns():
res = get_url('https://rest.uniprot.org/configure/uniprotkb/result-fields')
obj = json.loads(res.text)
result = {}
for i in range(len(obj)):
for col in obj[i]['fields']:
result[col['label']] = col['name']
return result
def get_id_mapping_fields():
res = get_url('https://rest.uniprot.org/configure/idmapping/fields')
obj = json.loads(res.text)
froms, tos = {}, {}
for group in range(len(obj['groups'])):
for item in obj['groups'][group]['items']:
if item['from']:
froms[item['displayName']] = item['name']
if item['to']:
tos[item['displayName']] = item['name']
return froms, tos
api_info = load_api_info()
columns_dict = get_uniprot_columns()
from_fields, to_fields = get_id_mapping_fields()
def get_arguments():
parser = ArgumentParser(description="UniProt Id Mapping through API",
epilog="A tool for retrieving information from UniProt.")
parser.add_argument(
"-i", "--input", help="""Input filename - can be:\n
\t1. a file containing a list of IDs (comma-separated values, no spaces)\n
\t2. a BLAST TSV result file (requires to be specified with the --blast parameter\n
\t3. a protein FASTA file to be annotated (requires the -db parameter)\n
\t4. nothing! If so, will read input from command line, and parse as CSV (id1,id2,...)""")
parser.add_argument("-o", "--output", help="Folder to store outputs", default="UPIMAPI_output")
parser.add_argument(
"-ot", "--output-table",
help="Filename of table output, where UniProt info is stored. If set, will override 'output' parameter "
"just for that specific file")
parser.add_argument(
"-rd", "--resources-directory", default=os.path.expanduser("~/upimapi_resources"),
help="Directory to store resources of UPIMAPI [~/upimapi_resources]")
parser.add_argument(
"-cols", "--columns", default=None, help="List of UniProt columns to obtain information from (separated by &)")
parser.add_argument(
"--from-db", default="UniProtKB AC/ID", choices=from_fields.keys(),
help="Which database are the IDs from. If from UniProt, default is fine [UniProtKB AC/ID]")
parser.add_argument(
"--to-db", default="UniProtKB", choices=to_fields.keys(),
help="To which database the IDs should be mapped. If only interested in columns information "
"(which include cross-references), default is fine [UniProtKB]")
parser.add_argument(
"--blast", action="store_true", default=False,
help="If input file is in BLAST TSV format (will consider one ID per line if not set) [false]")
parser.add_argument(
"--full-id", type=str2bool, default="auto", help="If IDs in database are in 'full' format: tr|XXX|XXX [auto]")
parser.add_argument(
"--fasta", help="Output will be generated in FASTA format [false]", action="store_true", default=False)
parser.add_argument(
"--step", type=int, default=1000, help="How many IDs to submit per request to the API [1000]")
parser.add_argument(
"--max-tries", default=3, type=int,
help="How many times to try obtaining information from UniProt before giving up [3]")
parser.add_argument("--sleep", default=3, type=int, help="Time between requests (in seconds) [3]")
parser.add_argument(
"--no-annotation", action="store_true", default=False,
help="Do not perform annotation - input must be in one of BLAST result or TXT IDs file or STDIN [false]")
parser.add_argument(
"--local-id-mapping", action="store_true", default=False,
help="Perform local ID mapping of SwissProt IDs. Advisable if many IDs of SwissProt are present [false]")
parser.add_argument(
"--skip-id-mapping", action="store_true", default=False,
help="If true, UPIMAPI will not perform ID mapping [false]")
parser.add_argument(
"--skip-id-checking", action="store_true", default=False,
help="If true, UPIMAPI will not check if IDs are valid before mapping [false]")
parser.add_argument(
"--skip-db-check", action="store_true", default=False,
help="DEPRECATED: This parameter does nothing now, and will be removed in future versions.")
parser.add_argument(
"--mirror", choices=['expasy', 'uniprot', 'ebi'], default='expasy',
help="From where to download UniProt database [expasy]")
parser.add_argument('-v', '--version', action='version', version=f'UPIMAPI {__version__}')
diamond_args = parser.add_argument_group('DIAMOND arguments')
diamond_args.add_argument(
"-db", "--database", default='uniprot',
help="How the reference database is inputted to UPIMAPI.\n"
"\t1. uniprot - UPIMAPI will download the entire UniProt and use it as reference\n"
"\t2. swissprot - UPIMAPI will download SwissProt and use it as reference\n"
"\t3. taxids - Reference proteomes will be downloaded for the taxa specified with the --taxids, and those "
"will be used as reference\n"
"\t4. a custom database - Input will be considered as the database, and will be used as reference")
diamond_args.add_argument(
"-t", "--threads", type=int, default=cpu_count(),
help="Number of threads to use in annotation steps [all available]")
diamond_args.add_argument(
"--evalue", type=float, default=1e-3, help="Maximum e-value to report annotations for [1e-3]")
diamond_args.add_argument(
"--pident", type=float, default=None, help="Minimum pident to report annotations for.")
diamond_args.add_argument(
"--bitscore", type=float, default=None, help="Minimum bit score to report annotations for (overrides e-value).")
diamond_args.add_argument(
"-mts", "--max-target-seqs", type=int, default=1,
help="Number of annotations to output per sequence inputed [1]")
diamond_args.add_argument(
"-b", "--block-size", type=float,
help="Billions of sequence letters to be processed at a time [memory / 20]")
diamond_args.add_argument(
"-c", "--index-chunks", type=int,
help="Number of chunks for processing the seed index [dependant on block size]")
diamond_args.add_argument(
"--max-memory", type=float, default=virtual_memory().available / (1024.0 ** 3),
help="Maximum memory to use (in Gb) [all available]")
diamond_args.add_argument(
"--taxids", default=None, help="Tax IDs to obtain protein sequences of for building a reference database.")
diamond_args.add_argument(
'--diamond-mode', help="Mode to run DIAMOND with [fast]", default='fast',
choices=['fast', 'mid_sensitive', 'sensitive', 'more_sensitive', 'very_sensitive', 'ultra_sensitive'])
special_functions = parser.add_argument_group('Special functions')
special_functions.add_argument(
"--show-available-fields", action="store_true", default=False,
help="Outputs the fields available from the API.")
args = parser.parse_args()
if args.show_available_fields:
sys.exit('\n'.join(columns_dict.keys()))
args.output = args.output.rstrip('/')
args.resources_directory = args.resources_directory.rstrip('/')
args.columns = args.columns.split('&') if args.columns else None
columns_fine = True
if args.columns:
for col in args.columns:
if col not in columns_dict.keys() and not col.startswith('Taxonomic lineage'):
print(
f'ERR: [{col}] is not a valid column name for ID mapping. For more information, check '
f'https://github.com/iquasere/UPIMAPI/tree/master#sometimes-the-return-fields-are-not-properly-updated')
columns_fine = False
if not columns_fine:
sys.exit(1)
if args.taxids is not None:
args.taxids = args.taxids.split(',')
return args
def timed_message(message):
print(f'{strftime("%Y-%m-%d %H:%M:%S", gmtime())}: {message}')
def human_time(seconds):
days = round(seconds // 86400)
if days > 0:
return strftime(f"{days}d%Hh%Mm%Ss", gmtime(seconds))
return strftime("%Hh%Mm%Ss", gmtime(seconds))
def str2bool(v):
if v.lower() == 'auto':
return 'auto'
elif v.lower() in ('yes', 'true', 't', 'y', '1'):
return True
elif v.lower() in ('no', 'false', 'f', 'n', '0'):
return False
else:
raise ArgumentTypeError('Boolean value expected.')
def get_fasta_ids(filename):
return [line[1:-1] for line in open(filename) if line.startswith('>')]
def parse_blast(blast):
result = pd.read_csv(blast, sep='\t', header=None)
result.columns = [
'qseqid', 'sseqid', 'pident', 'length', 'mismatch', 'gapopen', 'qstart', 'qend', 'sstart', 'send', 'evalue',
'bitscore']
return result
def string4mapping(columns=None):
if columns is None or columns == []: # if no columns are inputted, UPIMAPI uses defaults
return None
valid_columns = [column for column in columns if column in columns_dict.keys()]
invalid_columns = [column for column in columns if column not in columns_dict.keys()]
for col in invalid_columns:
print(f'WARNING: "{col}" is not a valid column name. '
f'Check https://www.uniprot.org/help/return_fields (Label* column) for valid column names '
f'or raise an issue at https://github.com/iquasere/UPIMAPI/issues')
return ','.join([columns_dict[column] for column in valid_columns])
def parallelize(data, func, num_of_processes=8):
data_split = np.array_split(data, num_of_processes)
pool = Pool(num_of_processes)
data = pd.concat(pool.map(func, data_split))
pool.close()
pool.join()
return data
def run_on_subset(func, data_subset, **kwargs):
return data_subset.apply(func, kwargs)
def parallelize_on_rows(data, func, num_of_processes=8, **kwargs):
return parallelize(data, partial(run_on_subset, func, kwargs), num_of_processes)
def uniprot_request(ids, columns=None, output_format='tsv'):
"""
Input:
ids: list of UniProt IDs to query
original_database: database from where the IDs are
database_destination: database to where to map (so far, only works with 'ACC'
output_format: format of response to get
columns: names of UniProt columns to get info on
databases: names of databases to cross-reference with
Output:
Returns the content of the response from UniProt
"""
fields = f'&fields={string4mapping(columns=columns)}' if output_format == 'tsv' else ''
WEBSITE_API = api_info['servers'][0]['url']
resp = get_url(f"{WEBSITE_API}/uniprotkb/accessions?accessions={','.join(ids)}{fields}&format={output_format}")
return resp.text
def split(a, n):
k, m = divmod(len(a), n)
return (a[i * k + min(i, m):(i + 1) * k + min(i + 1, m)] for i in range(n))
def submit_id_mapping(from_db, to_db, ids):
"""
Get info from one database to the other
:param from_db: Available options at https://rest.uniprot.org/configure/idmapping/fields
:param to_db: Available options at https://rest.uniprot.org/configure/idmapping/fields
:param ids:
:return:
"""
data = {"from": from_fields[from_db], "to": to_fields[to_db], "ids": ids}
r = requests.post(f"{api_info['servers'][0]['url']}/idmapping/run", data=data)
r.raise_for_status()
return r.json()["jobId"]
def get_id_mapping_results(job_id):
while True:
r = get_url(f"{api_info['servers'][0]['url']}/idmapping/status/{job_id}")
job = r.json()
if "jobStatus" in job:
if job["jobStatus"] == "RUNNING":
sleep(3)
else:
return r
def basic_idmapping(ids, from_db, to_db):
"""
Get info from one database to the other
:param ids:
:param from_db: Available options at https://rest.uniprot.org/configure/idmapping/fields
:param to_db: Available options at https://rest.uniprot.org/configure/idmapping/fields
:return:
"""
job_id = submit_id_mapping(from_db, to_db, ids)
r = get_id_mapping_results(job_id)
result = pd.DataFrame().from_dict(r.json()["results"])
while r.links.get("next", {}).get("url"):
r = get_url(r.links["next"]["url"])
result = pd.concat([result, pd.DataFrame().from_dict(r.json()["results"])])
return result
def basic_idmapping_batch(ids, from_db, to_db, step=1000):
"""
Allows to retrieve millions of IDs at once, there seems to be some limit causing UniProt's API to fail with
"Request Entity Too Large for url".
:param to_db:
:param from_db:
:param step:
:param ids:
:return:
"""
result = pd.DataFrame()
for i in tqdm(range(0, len(ids), step), desc='Getting valid UniProt IDs', ascii=' >='):
done = False
while not done:
j = min(i + step, len(ids))
try:
result = pd.concat([result, basic_idmapping(ids[i:j], from_db, to_db)])
done = True
except:
sleep(3)
return result
def basic_idmapping_multiprocess(ids, output, from_db, to_db, step=1000, threads=15):
result = pd.DataFrame()
ids_groups = split(ids, threads)
with Manager() as m:
with m.Pool() as p:
mapping_results = p.starmap(basic_idmapping_batch, [(
ids_group, from_db, to_db, step) for ids_group in ids_groups])
for res in mapping_results:
result = pd.concat([result, res])
timed_message(f'{result["from"].unique().size} IDs were successfully mapped.')
result.to_csv(output, sep='\t', index=False)
timed_message(f'Results saved at {output}.')
def get_valid_entries(ids):
job_id = submit_id_mapping("UniProtKB AC/ID", "UniProtKB", ids)
r = get_id_mapping_results(job_id)
valid_entries = [res["from"] for res in r.json()["results"] if '_' not in res["from"]]
while r.links.get("next", {}).get("url"):
r = get_url(r.links["next"]["url"])
valid_entries += [res["from"] for res in r.json()["results"] if '_' not in res["from"]]
return valid_entries
def get_valid_entries_batch(ids, step=1000):
"""
Allows to retrieve millions of IDs at once, there seems to be some limit causing UniProt's API to fail with
"Request Entity Too Large for url".
:param step:
:param ids:
:return:
"""
valid_entries = []
for i in tqdm(range(0, len(ids), step), desc='Getting valid UniProt IDs', ascii=' >='):
done = False
while not done:
j = min(i + step, len(ids))
try:
valid_entries += get_valid_entries(ids[i:j])
done = True
except:
sleep(3)
return valid_entries
def get_valid_entries_multiprocess(ids, step=1000, threads=15):
valid_entries = []
ids_groups = split(ids, threads)
with Manager() as m:
with m.Pool() as p:
result = p.starmap(get_valid_entries_batch, [(ids_group, step) for ids_group in ids_groups])
for res in result:
valid_entries += res
not_valid = [ide for ide in ids if ide not in valid_entries]
# take, from the valid IDs, the part after the dot, as this invalidates them
valid_entries = [entry.split('.')[0] for entry in valid_entries]
timed_message(f'{len(valid_entries)} UniProt IDs identified as valid.')
return valid_entries, not_valid
def get_uniprot_information(ids, step=1000, sleep_time=30, columns=None, max_tries=3):
"""
Input:
ids: list of UniProt IDs to query
step: INT, number of IDs to send per request
sleep_time: INT, number of seconds to wait between requests
columns: list - names of UniProt columns to get info on
Output:
pd.DataFrame will be returned with the information about the IDs queried.
"""
result = pd.DataFrame()
for i in tqdm(range(0, len(ids), step), desc=f'Retrieving UniProt information from {len(ids)} IDs'):
tries = 0
done = False
j = min(i + step, len(ids))
while not done and tries < max_tries:
try:
data = uniprot_request(ids[i:j], columns=columns)
if len(data) > 0:
uniprotinfo = pd.read_csv(StringIO(data), sep='\t')
result = pd.concat([result, uniprotinfo[uniprotinfo.columns.tolist()]])
sleep(sleep_time)
done = True
except ConnectionError:
print(f'ID mapping failed. Remaining tries: {max_tries - tries}')
tries += 1
sleep(10)
return result
def get_uniprot_fasta(ids, step=1000, sleep_time=30):
"""
Input:
ids: list of UniProt IDs to query
step: INT, number of IDs to send per request
sleep_time: INT, number of seconds to wait between requests
Output:
str object containing the fasta sequences and headers
of the proteis belonging to the IDs queried will be returned
"""
result = ''
for i in tqdm(range(0, len(ids), step), desc=f"Building FASTA from {len(ids)} IDs."):
j = min(i + step, len(ids))
data = uniprot_request(ids[i:j], output_format='fasta')
if len(data) > 0:
result += data
sleep(sleep_time)
return result
def uniprot_fasta_workflow(all_ids, output, max_iter=5, step=1000, sleep_time=10):
if os.path.isfile(output):
print(f'{output} was found. Will perform mapping for the remaining IDs.')
ids_done = get_fasta_ids(output)
else:
print(f'{output} not found. Will perform mapping for all IDs.')
ids_done = []
ids_missing = list(set(all_ids) - set(ids_done))
tries = 0
ids_done = ([ide.split('|')[1] for ide in get_fasta_ids(output)] if os.path.isfile(output) else [])
while len(ids_done) < len(all_ids) and tries < max_iter:
ids_missing = list(set([ide for ide in tqdm(all_ids, desc='Checking which IDs are missing information.')
if ide not in ids_done]))
print(f'Information already gathered for {int(len(ids_done) / 2)} ids. Still missing for {len(ids_missing)}.')
uniprotinfo = get_uniprot_fasta(ids_missing, step=step, sleep_time=sleep_time)
with open(output, 'a') as file:
file.write(uniprotinfo)
ids_done = [ide.split('|')[1] for ide in get_fasta_ids(output)]
tries += 1
if len(ids_done) == len(all_ids):
print(f'Results for all IDs are available at {output}')
else:
ids_unmapped_output = f"{'/'.join(output.split('/')[:-1])}/ids_unmapped.txt"
handler = open(ids_unmapped_output, 'w')
handler.write('\n'.join(ids_missing))
print(f'Maximum iterations were made. Results related to {str(len(ids_missing))} IDs were not obtained. '
f'IDs with missing information are available at {ids_unmapped_output} and information obtained is '
f'available at {output}')
def check_ids_already_done(output, ids):
if os.path.isfile(output) and os.stat(output).st_size > 1:
try:
result = pd.read_csv(output, sep='\t', low_memory=False).drop_duplicates()
print(f'{output} was found. Will perform mapping for the remaining IDs.')
if 'Entry Name' not in result.columns:
result['Entry Name'] = [np.nan] * len(result)
ids_done = result['Entry'].unique().tolist() + result['Entry Name'].unique().tolist()
except OSError: # file doesn't exist or is empty
print(f'{output} was found. However, it could not be parsed. Will restart mapping.')
result = pd.DataFrame()
ids_done = []
else:
print(f'{output} not found or empty. Will perform mapping for all IDs.')
result = pd.DataFrame()
ids_done = []
ids_missing = list(set(ids) - set(ids_done))
print(f'IDs present in uniprotinfo file: {int(len(ids_done) / 2)}') # entry and entry name count by 2
print(f'IDs missing: {len(ids_missing)}')
return ids_done, ids_missing, result
def select_columns(columns):
"""
:param columns: list - of columns to retrieve information from, including taxonomic columns
:return: new_cols: list - of columns to retrieve information from, without taxonomic columns added by UPIMAPI
:return: tax_cols: list - of taxonomic columns to retrieve information from
:return: taxids_cols: list - of taxid columns to retrieve information from
"""
if columns is None:
columns = [ # default columns of UPIMAPI
'Entry', 'Entry Name', 'Organism', 'Organism (ID)', 'Taxonomic lineage', 'Taxonomic lineage (Ids)',
'Gene Names', 'Protein names', 'EC number', 'Function [CC]', 'Pathway', 'Keywords',
'Protein existence', 'Gene Ontology (GO)', 'Protein families', 'BRENDA', 'BioCyc', 'CDD', 'eggNOG',
'Ensembl', 'InterPro', 'KEGG', 'Pfam', 'Reactome', 'RefSeq', 'UniPathway',
'Taxonomic lineage (SUPERKINGDOM)', 'Taxonomic lineage (PHYLUM)', 'Taxonomic lineage (CLASS)',
'Taxonomic lineage (ORDER)', 'Taxonomic lineage (FAMILY)', 'Taxonomic lineage (GENUS)',
'Taxonomic lineage (SPECIES)', 'Taxonomic lineage IDs (SPECIES)']
tax_cols = [col for col in columns if ('Taxonomic lineage (' in col and col not in [
'Taxonomic lineage (SPECIES)', 'Taxonomic lineage (Ids)'])]
taxids_cols = [col for col in columns if (
'Taxonomic lineage IDs (' in col and col not in 'Taxonomic lineage IDs (SPECIES)')]
for col in ['Entry Name', 'Entry']:
if col not in columns:
columns.insert(0, col)
new_cols = [col for col in columns if col not in tax_cols + taxids_cols + [
'Taxonomic lineage (SPECIES)', 'Taxonomic lineage IDs (SPECIES)']]
col_conversion = {'Organism': 'Taxonomic lineage (SPECIES)', 'Organism (ID)': 'Taxonomic lineage IDs (SPECIES)'}
for k, v in col_conversion.items():
if v in columns and k not in new_cols:
new_cols.append(k)
conditions = {
'len(tax_cols) > 0 and "Taxonomic lineage" not in new_cols': 'Taxonomic lineage',
'len(taxids_cols) > 0 and "Taxonomic lineage (Ids)" not in new_cols': 'Taxonomic lineage (Ids)',
'"Taxonomic lineage (SPECIES)" in columns and "Organism" not in new_cols': 'Organism',
'"Taxonomic lineage IDs (SPECIES)" in columns and "Organism (ID)" not in new_cols': 'Organism (ID)'}
for cond, col in conditions.items(): # check if cond (key) is True, then append or not the col (value)
if eval(cond):
new_cols.append(col)
for col in ['Entry Name', 'Entry']: # UPIMAPI requires these two columns to be present
if col not in new_cols:
new_cols.insert(0, col)
return columns, new_cols, tax_cols, taxids_cols
def make_taxonomic_lineage_df(tax_lineage_col, prefix='Taxonomic lineage IDs'):
"""
Parses the taxonomic lineage column of the uniprotinfo dataframe and returns a dataframe with the taxonomic lineage
separated in columns.
:param tax_lineage_col: pd.Series with the taxonomic lineage column of the uniprotinfo dataframe
:param prefix: str, prefix to use for the columns of the new dataframe
:return: pd.DataFrame with the taxonomic lineage separated in columns
"""
# First, split records by ', '
split_regex = r"(?<=\)),"
result = pd.DataFrame.from_records(tax_lineage_col.apply(lambda x: re.split(split_regex, x)).apply(
# Then, split each record by ' ('
lambda x: [part[:-1].split(' (') for part in x]).apply(
# Finally, build dictionary with the taxonomic level as key and the taxonomy as value. ' ('.join avoids cases
# where the taxonomy has a '(' in it (e.g. 'Clostridium scindens (strain JCM 10418 / VPI 12708) (species)')
lambda x: {part[-1]: ' ('.join(part[:-1]) for part in x if part[-1] != 'no rank'}))
# Rename columns in old UniProt fashion
result.rename(columns={col: f'{prefix} ({col.upper()})' for col in result.columns}, inplace=True)
for col in result.columns:
result[col] = result[col].str.lstrip()
return result
def uniprot_information_workflow(ids, output, max_iter=5, columns=None, step=1000, sleep_time=10):
ids_done, ids_missing, result = check_ids_already_done(output, ids)
tries, last_ids_missing, ids_unmapped_output = 0, None, f"{'/'.join(output.split('/')[:-1])}/ids_unmapped.txt"
columns, new_cols, tax_cols, taxids_cols = select_columns(columns)
columns = new_cols if columns is None else columns
uniprotinfo = pd.DataFrame()
while len(ids_missing) > 0 and tries < max_iter and ids_missing != last_ids_missing:
print(f'Information already gathered for {int(len(ids_done) / 2)} ids. Still missing for {len(ids_missing)}.')
last_ids_missing = ids_missing
info = get_uniprot_information(
ids_missing, step=step, columns=new_cols, max_tries=max_iter, sleep_time=sleep_time)
info.reset_index(inplace=True, drop=True)
if len(info) > 0:
ids_done += info['Entry'].unique().tolist() + info['Entry Name'].unique().tolist()
uniprotinfo = pd.concat([uniprotinfo, info], ignore_index=True)
ids_missing = list(set(last_ids_missing) - set(ids_done))
if len(ids_missing) > 0:
if last_ids_missing == ids_missing:
print("Could not map additional IDs for this mapping. There were probably some outdated IDs. "
"For more questions, please contact through https://github.com/iquasere/UPIMAPI/issues")
else:
print('Failed to retrieve information for some IDs. Retrying request.')
tries += 1
if len(uniprotinfo) == 0:
return result
tax_df = pd.DataFrame()
if len(tax_cols) > 0:
tax_df = make_taxonomic_lineage_df(uniprotinfo['Taxonomic lineage'], prefix='Taxonomic lineage')
if len(taxids_cols) > 0:
tax_df = pd.concat([tax_df, make_taxonomic_lineage_df(
uniprotinfo['Taxonomic lineage (Ids)'], prefix='Taxonomic lineage IDs')], axis=1)
# rename columns to old UniProt fashion if those columns are to be outputted
# then remove the original columns if they are not to be outputted
col_conversion = {'Organism': 'Taxonomic lineage (SPECIES)', 'Organism (ID)': 'Taxonomic lineage IDs (SPECIES)'}
for k, v in col_conversion.items():
if v in columns:
uniprotinfo[v] = uniprotinfo[k]
if k not in columns:
del uniprotinfo[k]
tax_df_gut_cols = [col for col in tax_df.columns if col not in col_conversion.values()] # don't repeat columns that were added in the previous loop
uniprotinfo = pd.concat([uniprotinfo, tax_df[tax_df_gut_cols]], axis=1)
result = pd.concat([result, uniprotinfo[columns]], ignore_index=True)
if len(ids_missing) == 0:
print(f'Results for all IDs are available at {output}')
else:
open(ids_unmapped_output, 'w').write('\n'.join(ids_missing))
print(f"Maximum iterations were made. Results related to {str(len(ids_missing))} IDs were not obtained. "
f"IDs with missing information are available at {ids_unmapped_output} and information obtained is "
f"available at {output}")
return result
def determine_full_id(ids):
for ide in ids:
if '|' in ide:
return True
return False
def parse_fasta(file):
with open(file) as f:
lines = [line.rstrip('\n') for line in f]
i = 0
sequences = {}
while i < len(lines):
if lines[i].startswith('>'):
name = lines[i][1:]
sequences[name] = ''
i += 1
while i < len(lines) and not lines[i].startswith('>'):
sequences[name] += lines[i]
i += 1
return sequences
def get_ids(args_input, input_type, full_id='auto'):
if args_input.endswith(('.zip', '.tar', '.gz', '.bz2')):
exit('File seems to be compressed! If not, please change its extension.')
if input_type == 'blast':
ids = parse_blast(args_input)['sseqid'].tolist()
elif input_type == 'txt':
ids = []
with open(args_input) as f:
preids = f.read().split('\n')
for preid in preids:
ids += preid.split(',')
elif input_type == 'fasta':
ids = parse_fasta(args_input).keys()
else: # if PIPE
ids = args_input.split(',')
if full_id == 'auto':
full_id = determine_full_id(ids)
print(f'Auto determined "full id" as: {full_id}')
if full_id:
return_ids = [ide.split('|')[1] for ide in ids if ide not in ['*', '']]
sp_ids = [ide.split('|')[1] for ide in ids if ide.startswith('sp')]
return return_ids, full_id, sp_ids
return_ids = [ide for ide in ids if ide not in ['*', '']]
return return_ids, full_id, return_ids
# second return_ids is just mock to return the same type of output as for sp_ids
def run_command(bash_command, print_message=True):
if print_message:
print(bash_command)
run(bash_command.split(), check=True)
def run_pipe_command(bash_command, output='', mode='w', print_message=True):
if print_message:
print(bash_command)
if output == '':
Popen(bash_command, stdin=PIPE, shell=True).communicate()
elif output == 'PIPE':
return Popen(bash_command, stdin=PIPE, shell=True,
stdout=PIPE).communicate()[0].decode('utf8')
else:
with open(output, mode) as output_file:
Popen(bash_command, stdin=PIPE, shell=True, stdout=output_file).communicate()
def make_diamond_database(fasta, dmnd):
run_command(f'diamond makedb --in {fasta} -d {dmnd}')
def block_size_and_index_chunks(argsb, argsc, memory):
if argsb:
b = argsb
else:
b = memory / 20 # b = memory in Gb / 20
if argsc:
return b, argsc
if b > 3:
return b, 1
if b > 2:
return b, 2
if b > 1:
return b, 3
return b, 4
def run_diamond(query, aligned, unaligned, database, threads=12, max_target_seqs=50, b=1, c=4, e_value=0.01,
bit_score=None, pident=None, mode='fast'):
command = (
f"diamond blastp --query {query} --out {aligned} --un {unaligned} --db {database} --outfmt 6 --unal 1 "
f"--threads {threads} --max-target-seqs {max_target_seqs} -b {b} -c {c} --evalue {e_value} "
f"--{mode.replace('_', '-')}")
if bit_score:
command += f' --min-score {bit_score}'
if pident:
command += f' --id {pident}'
run_command(command)
def get_proteome_for_taxid_slow(taxid, max_tries=3):
"""
Get proteome for taxid the "proper" way. It is very slow, though, so not used.
:param taxid:
:param max_tries:
:return:
"""
tries = 0
res = requests.get(f'https://rest.uniprot.org/uniprotkb/search?format=fasta&query=%28taxonomy_id%3A{taxid}%29')
result = res.content.decode('utf8')
pages = 0
while tries < max_tries and res.links != {}:
try:
res = requests.get(res.links['next']['url'])
result += res.content.decode('utf8')
pages += 1
except:
tries += 1
sleep(10)
return result
def get_proteome_for_taxid(taxid, max_tries=3):
tries = 0
url = f'https://rest.uniprot.org/uniprotkb/stream?format=fasta&query=taxonomy_id:{taxid}'
while tries < max_tries:
try:
return requests.get(url).content.decode('utf8')
except:
print(f'Failed! {max_tries - tries} tries remaining.')
tries += 1
sleep(10)
def local_uniprot_is_outdated(local_reldate_file):
local = open(local_reldate_file).readlines()
[sp_date, tr_date] = [datetime.strptime(local[i][:-1].split()[-1], '%d-%b-%Y') for i in [1, 2]]
current = requests.get("https://ftp.uniprot.org/pub/databases/uniprot/knowledgebase/complete/reldate.txt"
).content.decode('utf8').split('\n')
[c_sp_date, c_tr_date] = [datetime.strptime(current[i][:-1].split()[-1], '%d-%b-%Y') for i in [1, 2]]
return c_sp_date < sp_date or c_tr_date < tr_date
def download_with_progress_bar(url, output_folder):
# Streaming, so we can iterate over the response.
response = requests.get(url, stream=True)
total_size_in_bytes = int(response.headers.get('content-length', 0))
block_size = 102400 # 100 Kibibytes
progress_bar = tqdm(total=total_size_in_bytes, unit='iB', unit_scale=True, desc=f'Downloading {url.split("/")[-1]}')
with open(f'{output_folder}/{url.split("/")[-1]}', 'wb') as file:
for data in response.iter_content(block_size):
progress_bar.update(len(data))
file.write(data)
progress_bar.close()
def download_uniprot(output_folder, mirror='expasy'):
base_urls = {
'expasy': 'https://ftp.expasy.org',
'uniprot': 'https://ftp.uniprot.org/pub',
'ebi': 'https://ftp.ebi.ac.uk/pub'
}
for file in ["uniprot_sprot.fasta.gz", "uniprot_trembl.fasta.gz", "reldate.txt"]:
print(f'Downloading and writing: {file}')
download_with_progress_bar(
f'{base_urls[mirror]}/databases/uniprot/current_release/knowledgebase/complete/{file}', output_folder)
run_pipe_command(f'zcat {output_folder}/uniprot_trembl.fasta.gz {output_folder}/uniprot_sprot.fasta.gz > '
f'{output_folder}/uniprot.fasta')
for file in [f'{output_folder}/uniprot_trembl.fasta.gz', f'{output_folder}/uniprot_sprot.fasta.gz']:
os.remove(file)
def download_fasta_database(database, output_folder, taxids=None, max_tries=3, mirror='expasy'):
if database == 'uniprot':
download_uniprot(output_folder, mirror=mirror)
elif database == 'swissprot':
download_with_progress_bar(
"https://ftp.uniprot.org/pub/databases/uniprot/knowledgebase/complete/uniprot_sprot.fasta.gz",
output_folder)
run_command(f'gunzip {output_folder}/uniprot_sprot.fasta.gz')
elif database == 'taxids':
for taxid in tqdm(taxids, desc=f'Retrieving reference proteomes for {len(taxids)} taxa from UniProt.'):
with open(f'{output_folder}/taxids_database.fasta', 'a') as f:
f.write(get_proteome_for_taxid(taxid, max_tries=max_tries))
def must_download_database(database, db2suffix):
if database in db2suffix.keys():
return not os.path.isfile(db2suffix[database])
return False # custom database
def get_tabular_taxonomy(output):
res = requests.get('https://ftp.uniprot.org/pub/databases/uniprot/current_release/rdf/taxonomy.rdf.xz')
with open('taxonomy.rdf.xz', 'wb') as f:
f.write(res.content)
run_command(f'unxz taxonomy.rdf.xz')
print('Reading RDF taxonomy')
root = ET.parse('taxonomy.rdf').getroot()
elems = root.findall('{http://www.w3.org/1999/02/22-rdf-syntax-ns#}Description')
with open(output, 'w') as f:
written = f.write('\t'.join(['taxid', 'name', 'rank', 'parent_taxid']) + '\n')
for elem in tqdm(elems, desc='Converting XML taxonomy.rdf to TSV format'):
info = [elem.get('{http://www.w3.org/1999/02/22-rdf-syntax-ns#}about').split('/')[-1]]
scientific_name = elem.find('{http://purl.uniprot.org/core/}scientificName')
info.append(scientific_name.text if scientific_name is not None else '')
rank_elem = elem.find('{http://purl.uniprot.org/core/}rank')
info.append(rank_elem.get('{http://www.w3.org/1999/02/22-rdf-syntax-ns#}resource').split('/')[-1]
if rank_elem is not None else '')
upper_taxon = elem.find('{http://www.w3.org/2000/01/rdf-schema#}subClassOf')
info.append(upper_taxon.get('{http://www.w3.org/1999/02/22-rdf-syntax-ns#}resource').split('/')[-1]
if upper_taxon is not None else '')
written = f.write('\t'.join(info) + '\n')
def get_match_id(record, ids):
if record.entry_name in ids:
return record.entry_name
if record.accessions[0] in ids:
return record.accessions[0]
return None
def count_on_file(expression, file, compressed=False):
return int(check_output(f"{'zgrep' if compressed else 'grep'} -c '{expression}' {file}", shell=True))
def get_local_swissprot_data(sp_dat_filename, ids):
sp_dat = SP.parse(open(sp_dat_filename))
result, ids_found = [], []
i = 1
record = next(sp_dat)
number_of_entries = count_on_file('Reviewed;', sp_dat_filename)
while record is not None and len(ids) > 0:
match_id = get_match_id(record, ids)
if match_id is not None:
result.append(record.__dict__)
ids_found.append(match_id)
if i % 100000 == 0:
print(f'[{i}/{number_of_entries}] SwissProt entries queried')
record = next(sp_dat, None)
i += 1
print(f'[{i}/{number_of_entries}] SwissProt entries queried')
return pd.DataFrame(result), ids_found
def lineage_to_columns(lineage, tax_tsv):
l2c_result, l2c_taxids = {}, {}
for taxon in lineage:
match = tax_tsv.loc[taxon, ["rank", "taxid"]]
if type(match) == pd.core.series.Series:
rank, taxid = match[["rank", "taxid"]]
if type(rank) == str:
l2c_result[f'Taxonomic lineage ({rank.upper()})'] = taxon
l2c_taxids[f'Taxonomic identifier ({rank.upper()})'] = taxid
else: # some taxIDs have multiple levels (e.g. "Craniata")
for i in range(len(match)):
rank, taxid = match.iloc[i][["rank", "taxid"]]
if type(rank) == str:
l2c_result[f'Taxonomic lineage ({rank.upper()})'] = taxon
l2c_taxids[f'Taxonomic identifier ({rank.upper()})'] = taxid
l2c_result['Taxonomic lineage (ALL)'] = ', '.join(set(l2c_result.values()))
l2c_taxids['Taxonomic identifier (ALL)'] = ', '.join(set(l2c_taxids.values()))
l2c_result = {**l2c_result, **l2c_taxids, 'index': lineage}
return l2c_result
def lineages_to_columns(lineages, tax_tsv):
"""
Does the same as lineage_to_columns, but to all lineages, instead of a single one
:param lineages:
:param tax_tsv:
:return:
"""
return [lineage_to_columns(lineage, tax_tsv) for lineage in lineages]
def get_upper_taxids(taxid, tax_df):
"""
:param taxid: str - taxID to get upper taxIDs from
:param tax_df: pd.DataFrame - of read taxonomy.tsv (from taxonomy.rdf)
:returns list - of upper taxIDs
"""
if taxid == '0':
return []
taxids = []
while taxid != '1' and taxid != 'Taxon':
taxids.append(taxid)
taxid = tax_df.loc[taxid]['parent_taxid']
return taxids
def parse_taxonomy(data, tax_tsv_df, threads=15):
tax_tsv_df.set_index('name', inplace=True)
tax_tsv_df['taxid'] = tax_tsv_df['taxid'].astype(str)
all_classifications = split(data['organism_classification'].drop_duplicates().tolist(), threads)
with Manager() as m:
with m.Pool() as p:
result = p.starmap(lineages_to_columns, [(classifications, tax_tsv_df)
for classifications in all_classifications])
decompacted = []
for res in result:
decompacted += res
return pd.DataFrame(decompacted).set_index('index')
def parse_comments(sp_data):
result = []
bpc_list = []
for comments in sp_data['comments']:
partial = {key: '' for key in [
'FUNCTION', 'SUBUNIT', 'INTERACTION', 'SUBCELLULAR LOCATION', 'ALTERNATIVE PRODUCTS', 'TISSUE SPECIFICITY',
'PTM', 'POLYMORPHISM', 'DISEASE', 'MISCELLANEOUS', 'SIMILARITY', 'CAUTION', 'SEQUENCE CAUTION',
'WEB RESOURCE', 'MASS SPECTROMETRY', 'RNA EDITING', 'CATALYTIC ACTIVITY', 'COFACTOR', 'ACTIVITY REGULATION',
'PATHWAY', 'DEVELOPMENTAL STAGE', 'INDUCTION', 'ALLERGEN', 'BIOTECHNOLOGY', 'DISRUPTION PHENOTYPE',
'PHARMACEUTICAL', 'TOXIC DOSE', 'DOMAIN']}
bpc_dict = {'Kinetic parameters': []}
for comment in comments:
comment = comment.split(': ')
if comment[0] in partial.keys():
partial[comment[0]] += f'{": ".join(comment)} '
else:
if comment[0] in ['BIOPHYSICOCHEMICAL PROPERTIES']:
if comment[1] not in bpc_dict.keys():
bpc_dict[comment[1]] = [f'{comment[0]}: {comment[1]}: {comment[2]}']
else:
bpc_dict[comment[1]].append(f'{comment[0]}: {comment[1]}: {comment[2]}')
else:
print(f'Comment still not implemented: [{comment[0]}]')
result.append(partial)
bpc_dict['Kinetics'] = bpc_dict.pop('Kinetic parameters')
bpc_list.append(bpc_dict)
result = pd.DataFrame(result, columns=[
'Function [CC]', 'Subunit structure [CC]', 'Interacts with', 'Subcellular location [CC]',
'Alternative products (isoforms)', 'Tissue specificity', 'Post-translational modification', 'Polymorphism',
'Involvement in disease', 'Miscellaneous [CC]', 'Sequence similarities', 'Caution', 'Sequence caution',
'Web resources', 'Mass spectrometry', 'RNA editing', 'Catalytic activity', 'Cofactor', 'Activity regulation',
'Pathway', 'Developmental stage', 'Induction', 'Allergenic properties', 'Biotechnological use',
'Disruption phenotype', 'Pharmaceutical use', 'Toxic dose', 'Domain [CC]'])
result['Erroneous gene model prediction'] = result['Sequence caution']
bpc_df = pd.DataFrame(bpc_list)
for col in bpc_df:
bpc_df[col] = bpc_df[col].apply(lambda x: '; '.join(x) if type(x) == list else x)
return pd.concat([result, bpc_df], axis=1)
def add_to_dict(dictionary, key, value):
if key in dictionary.keys():
dictionary[key] += value
else:
dictionary[key] = value
def cross_references_to_columns(cross_refs):
result = {}
go_dict = {}
go_rel = {'C': 'cellular component', 'F': 'molecular function', 'P': 'biological process'}
for ref in cross_refs:
if ref[0] == 'GO':
refie = ref[2].split(':')
add_to_dict(go_dict, f'Gene ontology ({go_rel[refie[0]]})', f'{refie[1]} [{ref[1]}]; ')
add_to_dict(go_dict, 'Gene ontology (GO)', f'{refie[1]} [{ref[1]}]; ')
add_to_dict(go_dict, 'Gene ontology IDs', f'{ref[1]}; ')
else:
if ref[0] == 'Proteomes':
value = f'{ref[1]}: {ref[2]}'
else:
value = f'{ref[1]};'
add_to_dict(result, ref[0], value)
return result, go_dict