-
Notifications
You must be signed in to change notification settings - Fork 0
/
ProteoGenDB.py
2161 lines (1742 loc) · 96.9 KB
/
ProteoGenDB.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
import os
import io
import re
import shutil
import sys
import time
import yaml
import warnings
import pandas as pd
import numpy as np
import multiprocessing
import requests
import logging
import json
import platform
import vcf
from pyarrow import feather
from requests.exceptions import ConnectionError
from argparse import ArgumentParser
from datetime import datetime
from colorlog import ColoredFormatter
from Bio import SeqIO
from Bio.Seq import Seq
from Bio.SeqRecord import SeqRecord
from time import sleep
from tqdm import tqdm
global start_time
LOG_LEVEL = logging.DEBUG
LOGFORMAT = "%(log_color)s%(asctime)s %(message)s%(reset)s"
LOGFORMAT_WEL = "%(log_color)s%(message)s%(reset)s"
stream = logging.StreamHandler()
stream.setLevel(LOG_LEVEL)
stream.setFormatter(ColoredFormatter(LOGFORMAT_WEL))
log = logging
log.getLogger().setLevel(logging.INFO)
log.getLogger("requests").setLevel(logging.WARNING)
log.getLogger().addHandler(stream)
# Suppress SettingWithCopyWarning & PerformanceWarning
pd.options.mode.chained_assignment = None
warnings.simplefilter(action='ignore', category=pd.errors.PerformanceWarning)
def print_welcome():
if platform.system() == 'Windows':
os.system("color")
log.info(r"""
_____ _ _____ _____ ____
| __ \ | | / ____| | __ \| _ \
| |__) | __ ___ | |_ ___ ___ | | __ ___ _ __ | | | | |_) |
| ___/ '__/ _ \| __/ _ \/ _ \| | |_ |/ _ \ '_ \| | | | _ <
| | | | | (_) | || __/ (_) | |__| | __/ | | | |__| | |_) |
|_| |_| \___/ \__\___|\___/ \_____|\___|_| |_|_____/|____/
""")
log.info(" Niko Pinter - https://github.com/npinter/ProteoGenDB")
log.info(" v1.5 \n")
def multi_process(func, input_df, unit, *args):
with multiprocessing.Manager() as manager:
df_status = manager.dict()
progress_bar_value = manager.dict()
# set number of parallel jobs
if config_yaml["num_jobs"] == 0:
num_jobs = multiprocessing.cpu_count()
else:
num_jobs = config_yaml["num_jobs"]
if config_yaml["num_job_chunks"] == 0 or config_yaml["num_job_chunks"] == num_jobs:
num_job_chunks = num_jobs
else:
num_job_chunks = config_yaml["num_job_chunks"]
# setup all jobs
df_max = 0
df_step = int(np.ceil(len(input_df) / num_jobs))
jobs = []
# append progress bar job
progress_bar_process = multiprocessing.Process(target=progress_bar,
args=(len(input_df), progress_bar_value, unit))
# append sliced df to job list
for job_id in range(0, num_jobs):
df_min = df_max
df_max += df_step
if df_max > len(input_df):
df_max = len(input_df)
if not df_min == df_max:
df_slice = input_df[df_min:df_max]
process = multiprocessing.Process(target=globals()[func],
args=(df_slice,
df_status,
job_id,
progress_bar_value,
args))
jobs.append(process)
progress_bar_process.start()
for cj in range(0, num_jobs, num_job_chunks):
chunk_jobs = jobs[cj:cj+num_job_chunks]
# start jobs
for j in chunk_jobs:
j.start()
jobs_done = False
jobs_status_dict = {}
while not jobs_done:
for j in chunk_jobs:
if j.is_alive():
jobs_status_dict[j.name] = True
else:
jobs_status_dict[j.name] = False
if not all(value for value in jobs_status_dict.values()):
jobs_done = True
# validate if all jobs are finished
for j in chunk_jobs:
j.join()
progress_bar_process.join()
# job outputs need to be sorted by job id
sorted_dict = {k: df_status[k] for k in sorted(df_status)}
df_out = pd.concat(sorted_dict.values(), ignore_index=True)
df_status.clear()
return df_out
def progress_bar(max_value, p_bar_val, unit):
for bar_val in tqdm(range(max_value),
unit=" {}".format(unit),
bar_format='{percentage:1.0f}%|{bar:10}{r_bar}'):
p_bar_val_sum = sum(p_bar_val.values())
if bar_val > p_bar_val_sum:
sleep(1)
def read_yaml(yaml_file):
with open(yaml_file) as yaml_stream:
return yaml.safe_load(yaml_stream)
def read_fasta(fasta_file, database):
# create fasta data frame
with open(fasta_file) as fasta:
ids = []
gids = []
seqs = []
name = []
desc = []
fasta_df_temp = pd.DataFrame()
if database == "uniprot":
for seq_record in SeqIO.parse(fasta, "fasta"):
ids.append([seq_record.id.split("|")[1] if "|" in seq_record.id else seq_record.id][0])
seqs.append(seq_record.seq)
name.append([seq_record.id.split("|")[2] if "|" in seq_record.id else seq_record.id][0])
desc.append(seq_record.description)
fasta_df_temp["Identifier"] = ids
fasta_df_temp["Sequence"] = seqs
fasta_df_temp["Name"] = name
fasta_df_temp["Description"] = desc
elif database == "galaxy":
pids = []
gids = []
gname = []
for seq_record in SeqIO.parse(fasta, "fasta"):
header_split = seq_record.description.split("|")
pids.append(header_split[1]) # Protein ID
gids.append(header_split[4]) # Gene ID
gname.append(header_split[5]) # Gene name
seqs.append(seq_record.seq) # SAAV protein sequence
desc.append(header_split[6]) # Description
# Get UniProtIDs from EnsemblIDs
log.info("Requesting UniProtIDs with ENSEMBL ProteinIDs..")
uniprot_lookup_df = get_uniprot_id(pids, ens_sub=True).rename({"FromID": "ProteinID",
"ToID": "UniProtID"},
axis=1)
# fill dataframe
fasta_df_temp["ProteinID"] = pids
fasta_df_temp["UniProtID"] = multi_process("process_uniprot_ids", pids, "ids", uniprot_lookup_df)
fasta_df_temp["GeneID"] = gids
fasta_df_temp["GeneName"] = gname
fasta_df_temp["Sequence"] = seqs
fasta_df_temp["Description"] = desc
# fill empty strings in UniProtID column
fasta_df_temp["UniProtID"] = fasta_df_temp["UniProtID"].replace("", np.nan)
# get missing UniProtIDs via the GeneID
log.info("Requesting UniProtIDs with ENSEMBL GeneIDs..")
fasta_df_temp_uniprot_mis = fasta_df_temp[fasta_df_temp["UniProtID"].isnull()]
uniprot_lookup_genes = get_uniprot_id(fasta_df_temp_uniprot_mis["GeneID"].tolist(),
fmt_from="Ensembl",
ens_sub=True).rename({"FromID": "GeneID",
"ToID": "UniProtID"},
axis=1)
uniprot_lookup_genes = uniprot_lookup_genes.drop_duplicates(subset=["GeneID"])
# lookup UniProtIDs from GeneID df
fasta_df_temp = fasta_df_temp.set_index("GeneID").combine_first(
uniprot_lookup_genes.set_index("GeneID")
).reset_index()
# fill empty strings in UniProtID column
fasta_df_temp["UniProtID"] = fasta_df_temp["UniProtID"].replace(np.nan, "NoUniID")
elif database == "ensembl_merge":
for seq_record in SeqIO.parse(fasta, "fasta"):
ids.append(seq_record.id)
gids.append([seq_record.description.split("gene:")[1].split(" ", 1)[0]
if "gene:" in seq_record.description
else ""][0])
seqs.append(seq_record.seq)
name.append(seq_record.name.split(".")[0])
fasta_df_temp["ProteinID"] = ids
fasta_df_temp["StableID"] = name
fasta_df_temp["GeneID"] = gids
fasta_df_temp["Sequence"] = seqs
elif database == "ensembl":
for seq_record in SeqIO.parse(fasta, "fasta"):
ids.append(seq_record.id)
name.append(seq_record.name.split(".")[0])
gids.append([seq_record.description.split()[1]
if "ENSP" in seq_record.description and seq_record.description != ""
else seq_record.description][0])
seqs.append(seq_record.seq)
fasta_df_temp["ProteinID"] = ids
fasta_df_temp["StableID"] = name
fasta_df_temp["GeneID"] = gids
fasta_df_temp["Sequence"] = seqs
return fasta_df_temp
def read_strelka_vcf(vcf_path):
strelka_df = pd.DataFrame(columns=["ProteinID", "UniProtID", "VariantPos", "GeneID"])
if not isinstance(vcf_path, list):
vcf_path = [vcf_path]
for vcf_file in vcf_path:
vcf_reader = vcf.Reader(open(vcf_file, 'r'))
for record in vcf_reader:
info = record.INFO
# check if the variant is exonic and either nonsynonymous_SNV or stopgain
if 'ExonicFunc.refGeneWithVer' in info:
exonic_func = info['ExonicFunc.refGeneWithVer'][0]
if exonic_func not in ['nonsynonymous_SNV', 'stopgain', 'nonframeshift_deletion']:
continue
else:
continue
# extract AAChange information
if 'AAChange.refGeneWithVer' in info:
aa_changes = info['AAChange.refGeneWithVer']
for aa_change in aa_changes:
parts = aa_change.split(':')
if len(parts) >= 5:
gene_id, protein_id, _, _, aa_pos = parts[:5]
protein_id = protein_id.split('.')[0] # Remove version number
aa_pos = aa_pos.split('.')[-1] # Get only the amino acid change
# replace X to * (stopgain)
aa_pos = aa_pos.replace("X", "*")
# replace del to - (deletion)
aa_pos = aa_pos.replace("del", "-")
strelka_df = pd.concat([strelka_df, pd.DataFrame({
"ProteinID": protein_id,
"VariantPos": [[aa_pos]],
"GeneID": gene_id
})], ignore_index=True)
# map RefSeq IDs to UniProt IDs
# not all RefSeq IDs can be mapped to UniProt IDs!
unique_protein_ids = strelka_df["ProteinID"].unique().tolist()
uniprot_mapping = get_uniprot_id(
unique_protein_ids,
fmt_from="RefSeq_Nucleotide",
split_str=" "
)
# create lookup dictionary
uniprot_dict = dict(zip(uniprot_mapping["FromID"], uniprot_mapping["ToID"]))
# add UniProt IDs to the dataframe
strelka_df["UniProtID"] = strelka_df["ProteinID"].map(uniprot_dict)
# fill NaN values with "NoUniID"
strelka_df["UniProtID"] = strelka_df["UniProtID"].fillna("NoUniID")
return strelka_df
def subset_fasta_db(fasta_db, exp_mat):
fasta_df_temp = read_fasta(fasta_db, "uniprot")
ex_mat_sep = pd.read_csv(exp_mat, sep=None, iterator=True, engine="python")
ex_mat_sep_det = ex_mat_sep._engine.data.dialect.delimiter
ex_mat_df_temp = pd.read_csv(exp_mat, sep=ex_mat_sep_det)
id_subset = ex_mat_df_temp.iloc[:, 0]
fasta_subset = fasta_df_temp[fasta_df_temp["Identifier"].isin(id_subset.tolist())]
return fasta_subset
def write_fasta(df, out_path, time_str, fasta_type, in_src):
with open(os.path.join(out_path, "{}_{}_{}.fasta".format(time_str, fasta_type, in_src)), "w") as fasta_file:
SeqIO.write(df, fasta_file, "fasta")
def read_tso(tso_path):
with open(tso_path, "r") as tso_input:
tso_split = tso_input.read().split("[Small Variants]")[1][3:]
tso_df = pd.read_csv(io.StringIO(tso_split), sep="\t")
# if only one entry with NA is present return empty dataframe with all columns
if tso_df["P-Dot Notation"].isna().all():
tso_df = pd.DataFrame(columns=["P-Dot Notation", "ProteinID", "VariantPos3", "VariantPos", "Sequence"])
return tso_df
# extract NM ID
tso_df["ProteinID"] = tso_df["P-Dot Notation"].str.extract(r"(NP_\d+\.\d+).*")
# extract p. notation
# substitution: Ala102Val |[A-z]{3}
# fs mutation: Ser378PhefsTer6 |[A-z]{8}\d+
# inframe insertion: Ala767_Val769dup |_[A-z]{3}\d+dup
# stop gained: Arg661Ter |Ter
# inframe deletion: Pro129del |del
tso_df["VariantPos3"] = tso_df["P-Dot Notation"].str.extract(r"p\.\(([A-z]{3}\d+(?:[A-z]{3}|Ter|del))\)")
aa_dict = {
'Cys': 'C', 'Asp': 'D', 'Ser': 'S', 'Gln': 'Q', 'Lys': 'K',
'Ile': 'I', 'Pro': 'P', 'Thr': 'T', 'Phe': 'F', 'Asn': 'N',
'Gly': 'G', 'His': 'H', 'Leu': 'L', 'Arg': 'R', 'Trp': 'W',
'Ala': 'A', 'Val': 'V', 'Glu': 'E', 'Tyr': 'Y', 'Met': 'M',
'Ter': '*', 'del': '-'
}
aa_regex = "|".join(aa_dict.keys())
# translate into one-letter code and change to list
tso_df["VariantPos"] = tso_df["VariantPos3"].str.replace(
aa_regex, lambda x: aa_dict[x.group()], regex=True
).apply(lambda x: [x])
# drop na
tso_df = tso_df[~tso_df["VariantPos3"].isna()].reset_index(drop=True)
# get protein sequence from NCBI as SeqRecord
tso_df["Sequence"] = multi_process("fetch_fasta", tso_df["ProteinID"].tolist(), "ids", "ncbi")
return tso_df
def get_uniprot_id(ids, fmt_from="Ensembl_Protein", fmt_to="UniProtKB", ens_sub=False, split_str="_"):
response = []
base_url = "https://rest.uniprot.org/idmapping"
entrez_ids = [entrez_id.split(split_str)[0] for entrez_id in ids]
ids_per_batch = 100000
# remove duplicates from list
entrez_ids = list(dict.fromkeys(entrez_ids))
# add ENSEMBL IDs current subversion
if ens_sub:
if entrez_ids[0][:4] == "ENSP":
ensembl_lookup = read_fasta(config_yaml["fasta_ensembl"], "ensembl").drop(
["Sequence", "GeneID"],
axis=1)
ensembl_lookup = ensembl_lookup[~ensembl_lookup["StableID"].duplicated(keep="last")]
entrez_ids_df = pd.DataFrame(entrez_ids, columns=["StableID"])
entrez_ids_match = entrez_ids_df.merge(ensembl_lookup,
on='StableID',
how='left')
entrez_ids_match = entrez_ids_match[~entrez_ids_match["ProteinID"].isnull()]
entrez_ids = entrez_ids_match["ProteinID"].tolist()
elif entrez_ids[0][:4] == "ENSG":
ensembl_lookup = read_fasta(config_yaml["fasta_ensembl"], "ensembl").drop(
["Sequence", "ProteinID", "StableID"],
axis=1)
ensembl_lookup["StableID"] = ensembl_lookup["GeneID"].str.split(".").str[0]
ensembl_lookup = ensembl_lookup[~ensembl_lookup["StableID"].duplicated(keep="last")]
entrez_ids_df = pd.DataFrame(entrez_ids, columns=["StableID"])
entrez_ids_match = entrez_ids_df.merge(ensembl_lookup,
on='StableID',
how='left')
entrez_ids_match = entrez_ids_match[~entrez_ids_match["GeneID"].isnull()]
entrez_ids = entrez_ids_match["GeneID"].tolist()
# send REST API request
for entrez_id in range(0, len(entrez_ids), ids_per_batch):
entrez_ids_tmp = ",".join(entrez_ids[entrez_id:entrez_id + ids_per_batch])
data = {
'from': fmt_from,
'to': fmt_to,
'ids': entrez_ids_tmp
}
key_error = True
while key_error:
try:
response.append(json.loads(requests.post("{}/run".format(base_url), data=data).text)["jobId"])
key_error = False
except KeyError:
log.error("UniProt connection error.. retry..")
sleep(10)
result_df = pd.DataFrame()
uni_data = None
# get REST API results
err_json = True
err_result = True
for uni_job in response:
while err_result:
while err_json:
try:
uni_data = json.loads(requests.get("{}/stream/{}".format(base_url, uni_job)).text)
err_json = False
except json.decoder.JSONDecodeError:
log.error("Request failed.. retry..")
sleep(1)
try:
result_df = pd.concat([result_df, pd.DataFrame(uni_data["results"])])
err_result = False
except KeyError:
log.error("Waiting for results..")
err_json = True
sleep(5)
result_df = result_df.rename(columns={"from": "FromID", "to": "ToID"}).drop_duplicates(subset=["FromID"],
keep="first")
log.info("{}/{} ({}%) ENSEMBL IDs were mapped to a UniProt ID".format(
len(result_df),
len(entrez_ids),
round(len(result_df)/(len(entrez_ids)/100), 2))
)
return result_df
def fetch_fasta(pids, df_status, jid, p_bar_val, *args):
base_url = None
prot_seq_rec = None
seq_records = []
mode = args[0][0]
if mode == "ncbi":
for i_pid, pid in enumerate(pids):
# ToDo: check if temp database entry exists if not get from NCBI and write to disk
base_url = "https://www.ncbi.nlm.nih.gov/search/api/download-sequence/?db=protein&id={pid}&filename={pid}"
prot_seq_rec = SeqIO.read(io.StringIO(requests.get(base_url.format(pid=pid)).text), "fasta").seq
seq_records.append(prot_seq_rec)
p_bar_val[jid] = i_pid
elif mode == "ncbi_NM":
ncbi_api_key = args[0][1]
efetch_url = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi"
for i_pid, pid in enumerate(pids):
ncbi_api_status = True
while ncbi_api_status:
try:
params = {
"db": "nucleotide",
"id": pid,
"rettype": "fasta_cds_aa",
"retmode": "xml",
"api_key": ncbi_api_key
}
response = requests.get(efetch_url, params=params)
# extract the translated sequence
translated_seq = ''.join(response.text.split("\n")[1:])
prot_seq_rec = Seq(translated_seq)
seq_records.append(prot_seq_rec)
# when response 429 (rate limit), 500 (server error), .. log error
if response.status_code != 200:
log.error(f"{pid}: Error fetching sequences from NCBI! - retry..")
else:
ncbi_api_status = False
# small delay when no API key to avoid rate limiting
if ncbi_api_key == "":
sleep(5)
else:
sleep(2)
except Exception as e:
log.error(f"Error fetching sequence for {pid}: {str(e)}")
p_bar_val[jid] = i_pid
elif mode == "uniprot":
pid_list = [pids[i:i + 100] for i in range(0, len(pids), 100)]
pid_num = 0
for i_pid, pids in enumerate(pid_list):
pids_str = ",".join(pids)
base_url = "https://www.ebi.ac.uk/proteins/api/proteins?offset=0&size=100&accession={pids}"
seq_response = requests.get(base_url.format(pids=pids_str),
headers={"Accept": "application/json"},
timeout=10).text
seq_json = json.loads(seq_response)
# sort seq_json by order in pids
seq_json = sorted(seq_json, key=lambda k: pids.index(k['accession']))
for seq in seq_json:
seq_header = ">sp|{}|{}\n".format(
seq["accession"],
seq["id"]
)
seq_str = seq["sequence"]["sequence"]
prot_seq_rec = SeqIO.read(io.StringIO(seq_header + seq_str), "fasta").seq
seq_records.append(prot_seq_rec)
pid_num += len(pid_list[i_pid])
p_bar_val[jid] = pid_num
seq_records_df = pd.Series(seq_records)
df_status[jid] = seq_records_df
def get_annotation_data(pids, df_status, jid, p_bar_val, *args):
api_name = args[0][0]
api_url = args[0][1]
annot_folder = args[0][2]
api_annot_path = os.path.join(annot_folder)
# load Feather storage
feather_filename = "{}_data.feather".format(api_name)
feather_path = os.path.join(api_annot_path, feather_filename)
feather_temp_filename = "{}_data_{}.temp.feather".format(api_name, jid)
feather_temp_path = os.path.join(api_annot_path, feather_temp_filename)
try:
data = pd.read_feather(feather_path, columns=["pid", "data"])
except FileNotFoundError:
data = pd.DataFrame(columns=["pid", "data"])
feather.write_feather(data, feather_path)
data_new = pd.DataFrame(columns=["pid", "data"])
result_data = []
feather_save = False
# get pid data from Feather or update it via API
for i_pid, pid in enumerate(pids):
if pid in data["pid"].values:
result = data[data['pid'] == pid]
result_data.append(result["data"].iloc[0])
else:
con_err = True
while con_err:
try:
api_data = requests.get(api_url.format(pid=pid)).json()
con_err = False
except ConnectionError:
log.error("UniProt REST API issue.. retry..")
sleep(1)
data_new = pd.concat([data_new, pd.DataFrame({'pid': [pid], 'data': [api_data]})], ignore_index=True)
result_data.append(api_data)
feather_save = True
p_bar_val[jid] = i_pid
result_data_df = pd.Series(result_data)
if feather_save:
feather.write_feather(data_new, feather_temp_path)
df_status[jid] = result_data_df
def get_variant_info(pids, df_status, jid, p_bar_val, *args):
var_df = pd.DataFrame()
cfg = args[0][0]
for i_pid, pid in enumerate(pids):
variant_data = pid
if "features" in variant_data:
if variant_data["features"] is not None:
for variant in variant_data["features"]:
if cfg["variant_reviewed_filter"]:
if "clinicalSignificances" not in variant:
continue
if variant["clinicalSignificances"] is None:
continue
if not any(source in variant["clinicalSignificances"][0]["sources"] for source in cfg["variant_reviewed_filter"]):
continue
if cfg["variant_evidences_cutoff"] > 0:
if "evidences" not in variant:
continue
if variant["evidences"] is None:
continue
if len(variant["evidences"]) < cfg["variant_evidences_cutoff"]:
continue
var_dict = {}
if "ftId" in variant:
var_dict["DL_ftID"] = variant["ftId"]
else:
var_dict["DL_ftID"] = None
var_dict["DL_UniProtID"] = pid["accession"]
if "cosmic curated" in str(variant["xrefs"]):
for xref in variant["xrefs"]:
if xref["name"] == "cosmic curated":
var_dict["DL_CosmicID"] = xref["id"]
break
else:
var_dict["DL_CosmicID"] = None
if "ClinVar" in str(variant["xrefs"]):
for xref in variant["xrefs"]:
if xref["name"] == "ClinVar":
var_dict["DL_ClinVarID"] = xref["id"]
break
else:
var_dict["DL_ClinVarID"] = None
if "mutatedType" in variant:
var_seq_mutatedtype = variant["mutatedType"]
else:
var_seq_mutatedtype = None
var_dict["DL_VariantPosStr"] = "{}{}{}".format([variant["wildType"] if "wildType" in variant else "?"][0],
variant["begin"],
var_seq_mutatedtype)
var_dict["DL_VariantPos"] = [var_dict["DL_VariantPosStr"]]
if variant["begin"] != variant["end"]:
var_dict["DL_VariantPosStr"] = ""
var_dict["DL_VariantPos"] = []
if "populationFrequencies" in variant:
if not variant["populationFrequencies"] is None:
for pf in variant["populationFrequencies"]:
if pf["populationName"] == "MAF":
var_dict["DL_MAF"] = pf["frequency"]
break
else:
var_dict["DL_MAF"] = None
var_dict["DL_sig_patho"] = np.nan
var_dict["DL_sig_likely_patho"] = np.nan
var_dict["DL_sig_likely_benign"] = np.nan
var_dict["DL_sig_benign"] = np.nan
var_dict["DL_sig_uncertain"] = np.nan
var_dict["DL_sig_conflict"] = np.nan
if "clinicalSignificances" in variant:
if not variant["clinicalSignificances"] is None:
for cs in variant["clinicalSignificances"]:
if cs["type"] == "Likely pathogenic":
var_dict["DL_sig_likely_patho"] = ";".join(cs["sources"])
elif cs["type"] == "Pathogenic":
var_dict["DL_sig_patho"] = ";".join(cs["sources"])
elif cs["type"] == "Variant of uncertain significance":
var_dict["DL_sig_uncertain"] = ";".join(cs["sources"])
elif cs["type"] == "Benign":
var_dict["DL_sig_benign"] = ";".join(cs["sources"])
elif cs["type"] == "Likely benign":
var_dict["DL_sig_likely_benign"] = ";".join(cs["sources"])
elif cs["type"] == "Conflicting interpretations of pathogenicity":
var_dict["DL_sig_conflict"] = ";".join(cs["sources"])
if "association" in variant:
if not variant["association"] is None:
var_dict["DL_disease_association"] = "1"
else:
var_dict["DL_disease_association"] = "0"
else:
var_dict["DL_disease_association"] = "0"
var_df = pd.concat([var_df, pd.DataFrame(pd.Series(var_dict)).transpose()])
p_bar_val[jid] = i_pid
df_status[jid] = var_df
def map_variant_info(dl_df, df_status, jid, p_bar_val, *args):
input_df = args[0][0]
min_pep_len = args[0][1]
max_pep_len = args[0][2]
columns = input_df.columns.union(dl_df.columns)
input_df_temp = pd.DataFrame(columns=columns)
for index, var in dl_df.iterrows():
prot_temp_df = input_df.loc[
(input_df.UniProtID.isin([var.DL_UniProtID]))]
var_temp_df = prot_temp_df.loc[(prot_temp_df.VariantPos.isin(var.DL_VariantPos))]
if len(var_temp_df) > 0 and var_temp_df.VarSeqCleave.values[0] is not None:
if min_pep_len <= len(var_temp_df.VarSeqCleave.values[0]) <= max_pep_len:
var_temp_df = pd.concat([var_temp_df.reset_index(drop=True).drop(
["VarSeq"], axis=1),
pd.DataFrame(var).transpose().reset_index(drop=True)],
axis=1
)
input_df_temp = pd.concat([input_df_temp, var_temp_df])
p_bar_val[jid] = index
df_status[jid] = input_df_temp
def annotate_variant_info(input_df, left_join, annot_data_path, min_pep_len, max_pep_len):
log.info(f"Collect disease information for {left_join}s..")
# drop Sequence (probably lowers RAM usage)
input_df.drop(["Sequence"], axis=1, inplace=True)
input_df = input_df.explode('VariantPos')
def get_var_seq_cleave(row):
pos = row['VariantPos']
cleave_dict = row['VarSeqCleave'][0]
return cleave_dict.get(pos, '')
# apply function to each row
input_df['VarSeq'] = input_df.apply(get_var_seq_cleave, axis=1)
input_df['VarSeqCleave'] = input_df.apply(get_var_seq_cleave, axis=1)
# drop duplicates
input_df.drop_duplicates(subset=["UniProtID", "VariantPos", "VarSeqCleave"], keep="first", inplace=True)
# load Feather storage
feather_filename = "{}_data.feather".format("ebi")
feather_path = os.path.join(annot_data_path, feather_filename)
# create annotation folder
if not os.path.exists(annot_data_path):
os.mkdir(annot_data_path)
try:
feather_data = pd.read_feather(feather_path)
except FileNotFoundError:
feather_data = pd.DataFrame(columns=["pid", "data"])
feather.write_feather(feather_data, feather_path)
annotate_data = multi_process("get_annotation_data",
input_df.UniProtID.drop_duplicates().to_list(),
left_join,
"ebi",
"https://www.ebi.ac.uk/proteins/api/variation/{pid}",
annot_data_path)
# concatenate requested variant data and combine with Feather database
temp_feather_files = []
for file_name in os.listdir(annot_data_path):
if file_name.endswith(".temp.feather"):
temp_feather_files.append(os.path.join(annot_data_path, file_name))
# save new annotations to Feather database
if len(temp_feather_files) > 0:
temp_feather_dfs = pd.DataFrame()
for temp_feather_file in temp_feather_files:
temp_feather_dfs = pd.concat([temp_feather_dfs, pd.read_feather(temp_feather_file)])
os.remove(temp_feather_file)
feather_data = pd.concat([feather_data, temp_feather_dfs])
feather.write_feather(feather_data, feather_path)
log.info(f"Extract variant information for {left_join}s..")
disease_lookup_df = multi_process("get_variant_info",
annotate_data,
left_join,
config_yaml)
# remove duplicates with UniProtID and VarPos
disease_lookup_df = disease_lookup_df.drop_duplicates(
subset=["DL_UniProtID", "DL_VariantPosStr"])
log.info(f"Map variant information..")
input_df = multi_process("map_variant_info",
disease_lookup_df,
"Variants",
input_df,
min_pep_len,
max_pep_len)
# keep only columns for FASTA header
input_df_drop_list = ["DL_UniProtID", "DL_CosmicID", "DL_VariantPos", "DL_VariantPosStr"]
input_df.drop(input_df_drop_list, axis=1, inplace=True)
# replace nan with NA
input_df.replace({np.nan: "NA"}, inplace=True)
# rename columns
input_df.rename(columns=lambda x: x.replace("DL_", ""), inplace=True)
return input_df
def process_uniprot_ids(ids, df_status, jid, p_bar_val, *args):
lookup_df = args[0][0].reset_index(drop=True)
uids = []
out_df = pd.DataFrame()
lookup_df["ProteinID"] = lookup_df.ProteinID.str[:15]
for i_pid, pid in enumerate(ids):
uid = lookup_df["UniProtID"].to_numpy()[lookup_df["ProteinID"].to_numpy() == pid[:15]]
uids.append([uid[0] if len(uid) > 0 else ""][0])
p_bar_val[jid] = i_pid
out_df["UniProtID"] = uids
df_status[jid] = out_df["UniProtID"]
def annotate_saavs(fasta_df, cfg):
log.info("Annotate SAAVs from Galaxy Workflow..")
# extract SAAVs position from ProteinID
fasta_df_temp = fasta_df
fasta_df_temp["VariantPos"] = fasta_df["ProteinID"].str.split(pat="_").str[-1]
# logic if ProteinID == VariantPos --> isoform/None
fasta_df_temp.loc[(fasta_df_temp.VariantPos == fasta_df_temp.ProteinID), 'VariantPos'] = None
# split iso and saavs
fasta_df_var = fasta_df_temp[fasta_df_temp.VariantPos.notnull()]
# fasta_df_iso = fasta_df_temp[fasta_df_temp.VariantPos.isnull()]
# multiple SAAVs to list
fasta_df_var["VariantPos"] = fasta_df_var["VariantPos"].str.split(pat=".")
# generate variant sequences
fasta_df_var = multi_process("process_saavs", fasta_df_var, "variants", cfg)
return fasta_df_var
def process_saavs(df, df_status, jid, p_bar_val, *args):
cfg = args[0][0]
df["VarSeq"] = None
df["VarSeqCleave"] = None
var_seq_len = cfg["var_seq_length"]
for pos in df.itertuples():
var_seq_temp = {}
var_seq_cleave_temp = {}
if pos.Sequence is not None:
for p in pos.VariantPos:
if not p[-1].isnumeric():
# substitution, termination, deletion
aa_pos = int(p[1:-1])-1
seq_pos_last = len(pos.Sequence)
pos_start = [aa_pos - var_seq_len if aa_pos - var_seq_len > 0 else 0][0]
pos_end = [aa_pos + var_seq_len + 1 if aa_pos + var_seq_len + 1 < seq_pos_last else seq_pos_last][0]
pre_x = ["#" * (2 * var_seq_len + 1 - (pos_end-pos_start))
if pos_start == 0 else ""][0]
post_x = ["#" * (2 * var_seq_len + 1 - (pos_end-pos_start))
if aa_pos + var_seq_len + 1 > seq_pos_last else ""][0]
var_seq_temp[p] = pre_x + pos.Sequence[pos_start:pos_end] + post_x
var_seq_cleave_temp[p] = cleave_sequence(var_seq_temp[p], cfg)
df["VarSeq"].loc[pos.Index] = [var_seq_temp]
df["VarSeqCleave"].loc[pos.Index] = [var_seq_cleave_temp]
else:
df["VarSeq"].loc[pos.Index] = [{}]
df["VarSeqCleave"].loc[pos.Index] = [{}]
# update progress bar value process-wise
p_bar_val[jid] = pos.Index
# drop everything with invalid var seq
df = df[df["VarSeq"].str[0] != {}]
df_status[jid] = df
def process_isoforms(df, df_status, jid, p_bar_val, *args):
cfg = args[0][0]
iso_seq_cleave_temp = {}
for iso in df.itertuples():
# cleave isoform
iso_seq_cleave_temp[iso.Identifier] = cleave_sequence(iso.Sequence, cfg, full_protein=True)
# update progress bar value process-wise
p_bar_val[jid] = iso.Index
# transform into dataframe
df_new = pd.Series(iso_seq_cleave_temp).to_frame().reset_index()
df_new.columns = ["UniProtID", "VarSeqCleave"]
if cfg["drop_unmapped_isoforms"]:
df_new = pd.merge(df_new, df[['Identifier', 'Consensus']],
left_on='UniProtID',
right_on='Identifier',
how='left')
df_new = df_new.drop('Identifier', axis=1)
df_status[jid] = df_new
def process_mutation(df, df_status, jid, p_bar_val, *args):
df["SequenceMut"] = None
for i_seq, seq in df.iterrows():
var_pos = int(seq.VariantPos[0][1:-1])
var_sub = seq.VariantPos[0][-1]
var_aa = seq.VariantPos[0][0]
# keep only variants where consensus AA occurs is in sequence
# --> missmatch of mapped sequence (ENSEMBL (read_fasta, "ensembl") or NCBI (fetch_fasta)
if len(seq.Sequence) < var_pos:
# drop sequence if variant position is out of range
df["SequenceMut"].loc[i_seq] = Seq("")
else:
if seq.Sequence[var_pos - 1:var_pos][0] == var_aa:
if not var_sub == "-" and not var_sub == "*":
df["SequenceMut"].loc[i_seq] = seq.Sequence[:var_pos - 1] + var_sub + seq.Sequence[var_pos:]
elif var_sub == "*":
df["SequenceMut"].loc[i_seq] = seq.Sequence[:var_pos - 1]
elif var_sub == "-":
df["SequenceMut"].loc[i_seq] = seq.Sequence[:var_pos - 1] + seq.Sequence[var_pos:]
else:
df["SequenceMut"].loc[i_seq] = Seq("")
# update progress bar
p_bar_val[jid] = i_seq
df = df.drop(columns=["Sequence"]).rename(columns={"SequenceMut": "Sequence"})
df_status[jid] = df
def cleave_sequence(var_dict, cfg, full_protein=False):
regex = re.compile(cfg["enzyme_specificity"])
enz_spec = [(str(var_dict).index(enz_match) + 1, str(var_dict).index(enz_match) + len(enz_match), enz_match)
for enz_match in regex.findall(str(var_dict))]
enz_peptide = None
cleaved_peptides = []
for e in enz_spec:
enz_pos_low, enz_pos_high, enz_group = e
if full_protein:
if cfg["min_spec_pep_len"] <= enz_pos_high - enz_pos_low + 1 <= cfg["max_spec_pep_len"]:
cleaved_peptides.append(Seq(enz_group))
else:
if enz_pos_low <= cfg["var_seq_length"] + 1 <= enz_pos_high:
# remove # from N- or C- terminal peptides
if "#" in enz_group:
enz_group = enz_group.replace("#", "")
enz_peptide = Seq(enz_group)
return enz_peptide
else:
enz_peptide = None
if not enz_spec:
enz_peptide = None
if full_protein:
return cleaved_peptides
return enz_peptide