Newer
Older
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
# input:
# the family PED file [${BATCH_ID}_${PLATE_ID}_${FAMILY_ID}.ped]
# the VCF folder containing the proband VCF [${PROJECT_ID}/VCF/ ${PLATE_ID}_${FAMILY_ID}.ready.${CHILD_ID}_${FAMILY_ID}.vcf.gz]
# G2P text output for the trio [${FAMILY_ID}.report.txt]
#
#
# output:
# DECIPHER formated file for the proband
# - all (filtered) G2P variants
#
# checks:
# all G2P variants found in the individual VCF
#
# Author: MH
# last modified: MARCH 04, 2022
import sys
import os
import csv
import gzip
from collections import defaultdict
ASSEMBLY = 'GRCh38'
INTERGENIC = 'No'
ACCESS = 'No'
G2P_DICT = {} # key: chr:pos:ref:alt; value: 0 (if found only in G2P); 1 (if found in VCF) - for variants found in G2P output for this CHILD_ID
G2P_DATA = {} # key: chr:pos:ref:alt; value: (transcript,gene,GT)
NUM_UNIQ_G2P_VARS = 0
CHILD_ID = 0
CHILD_SEX = 0
DEC_CHILD_SEX = 'unknown'
ALL_CHILD_DICT = {} # key: chr:pos:ref:alt; value: (num_ALT_reads,VAF)
CHILD_INHER_DICT = {} # key: chr:pos:ref:alt; value: "Unknown" - all variants in a singleton are of unknown inheritance
SNAP_FLANK = 25
MAP_DICT = {} # key: family_id (aka decipher_id); value: internal (decipher) ID
TRANS_DICT = {} # key: transcriptID not found in DECIPHER; value: the chosen replacement transcriptID from those available in DECIPHER
FS_THRESH = float(60)
SOR_THRESH = float(3)
def go(dec_map_file,trans_map_file,ped_file,in_g2p_file,fam_igv_dir,vcf_dir,plate_id,fam_id,dec_dir,fam_bam_dir):
# read the decipher to internal ID mapping file
read_map_file(dec_map_file)
# read the transcript mapping file
read_trans_map(trans_map_file)
# read the ped file and establish CHILD_ID,CHILD_SEX,MOM_ID,DAD_ID
read_ped(ped_file)
if (CHILD_ID != 0) and (CHILD_SEX != 0) and (DEC_CHILD_SEX != 'unknown'):
print "======================================"
print "Analyzing singleton CHILD_ID = %s, CHILD_SEX = %s, DEC_CHILD_SEX = %s" % (CHILD_ID,CHILD_SEX,DEC_CHILD_SEX)
print "======================================"
sys.stdout.flush()
else:
print "ERROR: problems reading the PED file = %s" % (ped_file)
raise SystemExit
# read the G2P output for this family
read_G2P(in_g2p_file)
# now read the proband VCFs and record all the variants
child_vcf_file = '%s/%s_%s.ready.%s.vcf.gz' % (vcf_dir,plate_id,fam_id,CHILD_ID)
read_all_VCF_vars(child_vcf_file,ALL_CHILD_DICT)
print "Found %s unique VCF variants for CHILD (%s)" % (len(ALL_CHILD_DICT),CHILD_ID)
sys.stdout.flush()
# now go over all child variants and set the inheritance to "Unknown"
num_child_vars_assigned = 0
for key,v in ALL_CHILD_DICT.iteritems():
CHILD_INHER_DICT[key] = 'Unknown'
assigned_ratio = (float(num_child_vars_assigned)/float(len(ALL_CHILD_DICT)))*100.0
print "%s of the %s unique VCF variants (%.2f%%) for CHILD (%s) has been assigned to parents" % (num_child_vars_assigned,len(ALL_CHILD_DICT),assigned_ratio,CHILD_ID)
sys.stdout.flush()
# setup the DECIPHER output file
out_dec_file = '%s/%s_DEC_FLT.csv' % (dec_dir,CHILD_ID) ################################
out_han = open(out_dec_file,'w')
out_han.write('Internal reference number or ID,Chromosome,Start,Genome assembly,Reference allele,Alternate allele,Transcript,Gene name,Intergenic,Chromosomal sex,Other rearrangements/aneuploidy,Open-access consent,Age at last clinical assessment,Prenatal age in weeks,Note,Inheritance,Pathogenicity,Phenotypes,HGVS code,Genotype,Responsible contact\n')
# setup the IGV snapshot file
out_igv_file = '%s/IGV/%s.solo.snapshot.FLT.txt' % (dec_dir,CHILD_ID) #################################
out_igv_han = open(out_igv_file,'w')
out_igv_han.write('new\n')
out_igv_han.write('genome hg38\n')
out_igv_han.write('mkdir -p "%s"\n' % (fam_igv_dir))
out_igv_han.write('new\n')
child_bam = '%s/%s/%s-ready.bam' % (fam_bam_dir,CHILD_ID,CHILD_ID)
out_igv_han.write('load %s\n' % (child_bam))
out_igv_han.write('snapshotDirectory "%s"\n' % (fam_igv_dir))
out_igv_han.write('\n')
# now read the child VCF, check if the variant in the G2P output, if yes:
# set the value in the dict to 1
# print out to to output file
in_cntr = 0
out_cntr = 0
child_vcf_file = '%s/%s_%s.ready.%s.vcf.gz' % (vcf_dir,plate_id,fam_id,CHILD_ID)
in_han = gzip.open(child_vcf_file,'r')
for line in in_han:
if line.startswith('#'):
continue
in_cntr += 1
data = [x.strip() for x in line.strip().split('\t')]
chr = data[0]
pos = int(data[1])
ref = data[3]
alt = data[4]
# extract FS and SOR
FS = ''
SOR = ''
infos = [y.strip() for y in data[7].strip().split(';')]
for info in infos:
if info.startswith('FS='):
tag,FS = info.split('=')
FS = float(FS)
elif info.startswith('SOR='):
tag,SOR = info.split('=')
SOR = float(SOR)
VCF_VAR = data[9]
key = '%s:%s:%s:%s' % (chr,pos,ref,alt)
inher_stat = CHILD_INHER_DICT[key]
##############################################################
# different processing depending on being a SNP, INS, or DEL #
##############################################################
if len(ref) == len(alt): # SNP
if len(ref) != 1:
print "ERROR: MNPs are not supported!"
print line
raise SystemExit
key_to_match = '%s:%s:%s:%s' % (chr,pos,ref,alt)
if key_to_match in G2P_DICT:
G2P_DICT[key_to_match] = 1
trans = G2P_DATA[key_to_match][0]
gene = G2P_DATA[key_to_match][1]
GT = G2P_DATA[key_to_match][2]
if (chr != 'chrX') and (chr != 'chrY'):
if GT == 'HET':
genotype = 'Heterozygous'
elif GT == 'HOM':
genotype = 'Homozygous'
else:
print "ERROR: Cannot understand GT = %s" % (GT)
raise SystemExit
elif (chr == 'chrX') or (chr == 'chrY'):
if DEC_CHILD_SEX == '46XX': # a girl
if GT == 'HET':
genotype = 'Heterozygous'
elif GT == 'HOM':
genotype = 'Homozygous'
else:
print "ERROR: Cannot understand GT = %s" % (GT)
raise SystemExit
elif DEC_CHILD_SEX == '46XY': # a boy
if GT == 'HET':
genotype = 'Heterozygous'
print " WARNING: HET variant on chrX/Y for a boy (%s): %s\t%s\t%s\t%s\t%s" % (CHILD_ID,chr,pos,ref,alt,VCF_VAR)
elif GT == 'HOM':
genotype = 'Hemizygous'
else:
print "ERROR: Cannot understand GT = %s" % (GT)
raise SystemExit
else:
print "ERROR: unknown sex for this proband = %s" % (DEC_CHILD_SEX)
raise SystemExit
else:
print "ERROR: unknown chr"
print line
raise SystemExit
# write to the DECIPHER file
gene_id_idx = gene.find('(')
if gene_id_idx == -1:
gene_id_idx = len(gene)
gene_id = gene[0:gene_id_idx]
int_ID = MAP_DICT[fam_id]
if trans in TRANS_DICT: # if the transcriptID is to be replaced
safe_trans = TRANS_DICT[trans]
else:
safe_trans = trans
to_write = '%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,,%s,,,,"%s",,,,%s,\n' % (int_ID,chr[3:],pos,ASSEMBLY,ref,alt,safe_trans,gene_id,INTERGENIC,DEC_CHILD_SEX,ACCESS,inher_stat,genotype)
out_cntr += 1
out_han.write(to_write)
# write to the IGV file
i_s = pos - SNAP_FLANK
i_e = pos + SNAP_FLANK
# check if above FS/SOR_THRESH to include in the snapshot name
if (FS == '') or (SOR == ''):
flag = 'NA'
elif (FS >= FS_THRESH) and (SOR >= SOR_THRESH):
flag = 'FS_%.1f_SOR_%.1f' % (FS,SOR)
else:
flag = 'OK'
i_name = '%s_%s_%s_%s_%s_%s.png' % (CHILD_ID,chr,pos,ref,alt,flag)
out_igv_han.write('goto %s:%s-%s\n' % (chr,i_s,i_e))
out_igv_han.write('sort strand\n')
out_igv_han.write('squish\n')
out_igv_han.write('snapshot %s\n' % (i_name))
out_igv_han.write('\n')
elif len(ref) > len(alt): # DEL
if len(alt) != 1:
print "ERROR with a deletion"
print line
raise SystemExit
G2P_key_to_match = '%s:%s:%s:-' % (chr,pos+1,ref[1:])
if G2P_key_to_match in G2P_DICT:
G2P_DICT[G2P_key_to_match] = 1
trans = G2P_DATA[G2P_key_to_match][0]
gene = G2P_DATA[G2P_key_to_match][1]
GT = G2P_DATA[G2P_key_to_match][2]
if (chr != 'chrX') and (chr != 'chrY'):
if GT == 'HET':
genotype = 'Heterozygous'
elif GT == 'HOM':
genotype = 'Homozygous'
else:
print "ERROR: Cannot understand GT = %s" % (GT)
raise SystemExit
elif (chr == 'chrX') or (chr == 'chrY'):
if DEC_CHILD_SEX == '46XX': # a girl
if GT == 'HET':
genotype = 'Heterozygous'
elif GT == 'HOM':
genotype = 'Homozygous'
else:
print "ERROR: Cannot understand GT = %s" % (GT)
raise SystemExit
elif DEC_CHILD_SEX == '46XY': # a boy
if GT == 'HET':
genotype = 'Heterozygous'
print " WARNING: HET variant on chrX/Y for a boy (%s): %s\t%s\t%s\t%s\t%s" % (CHILD_ID,chr,pos,ref,alt,VCF_VAR)
elif GT == 'HOM':
genotype = 'Hemizygous'
else:
print "ERROR: Cannot understand GT = %s" % (GT)
raise SystemExit
else:
print "ERROR: unknown sex for this proband = %s" % (DEC_CHILD_SEX)
raise SystemExit
else:
print "ERROR: unknown chr"
print line
raise SystemExit
# write to the DECIPHER file
gene_id_idx = gene.find('(')
if gene_id_idx == -1:
gene_id_idx = len(gene)
gene_id = gene[0:gene_id_idx]
int_ID = MAP_DICT[fam_id]
if trans in TRANS_DICT: # if the transcriptID is to be replaced
safe_trans = TRANS_DICT[trans]
else:
safe_trans = trans
to_write = '%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,,%s,,,,"%s",,,,%s,\n' % (int_ID,chr[3:],pos,ASSEMBLY,ref,alt,safe_trans,gene_id,INTERGENIC,DEC_CHILD_SEX,ACCESS,inher_stat,genotype)
out_cntr += 1
out_han.write(to_write)
# write to the IGV file
i_s = pos - SNAP_FLANK
i_e = pos + SNAP_FLANK
# check if above FS/SOR_THRESH to include in the snapshot name
if (FS == '') or (SOR == ''):
flag = 'NA'
elif (FS >= FS_THRESH) and (SOR >= SOR_THRESH):
flag = 'FS_%.1f_SOR_%.1f' % (FS,SOR)
else:
flag = 'OK'
i_name = '%s_%s_%s_%s_%s_%s.png' % (CHILD_ID,chr,pos,ref,alt,flag)
out_igv_han.write('goto %s:%s-%s\n' % (chr,i_s,i_e))
out_igv_han.write('sort strand\n')
out_igv_han.write('squish\n')
out_igv_han.write('snapshot %s\n' % (i_name))
out_igv_han.write('\n')
elif len(ref) < len(alt): # INS
if len(ref) != 1:
print "ERROR with an insertion"
print line
raise SystemExit
G2P_key_to_match = '%s:%s:-:%s' % (chr,pos+1,alt[1:])
if G2P_key_to_match in G2P_DICT:
G2P_DICT[G2P_key_to_match] = 1
trans = G2P_DATA[G2P_key_to_match][0]
gene = G2P_DATA[G2P_key_to_match][1]
GT = G2P_DATA[G2P_key_to_match][2]
if (chr != 'chrX') and (chr != 'chrY'):
if GT == 'HET':
genotype = 'Heterozygous'
elif GT == 'HOM':
genotype = 'Homozygous'
else:
print "ERROR: Cannot understand GT = %s" % (GT)
raise SystemExit
elif (chr == 'chrX') or (chr == 'chrY'):
if DEC_CHILD_SEX == '46XX': # a girl
if GT == 'HET':
genotype = 'Heterozygous'
elif GT == 'HOM':
genotype = 'Homozygous'
else:
print "ERROR: Cannot understand GT = %s" % (GT)
raise SystemExit
elif DEC_CHILD_SEX == '46XY': # a boy
if GT == 'HET':
genotype = 'Heterozygous'
print " WARNING: HET variant on chrX/Y for a boy (%s): %s\t%s\t%s\t%s\t%s" % (CHILD_ID,chr,pos,ref,alt,VCF_VAR)
elif GT == 'HOM':
genotype = 'Hemizygous'
else:
print "ERROR: Cannot understand GT = %s" % (GT)
raise SystemExit
else:
print "ERROR: unknown sex for this proband = %s" % (DEC_CHILD_SEX)
raise SystemExit
else:
print "ERROR: unknown chr"
print line
raise SystemExit
# write to the DECIPHER file
gene_id_idx = gene.find('(')
if gene_id_idx == -1:
gene_id_idx = len(gene)
gene_id = gene[0:gene_id_idx]
int_ID = MAP_DICT[fam_id]
if trans in TRANS_DICT: # if the transcriptID is to be replaced
safe_trans = TRANS_DICT[trans]
else:
safe_trans = trans
to_write = '%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,,%s,,,,"%s",,,,%s,\n' % (int_ID,chr[3:],pos,ASSEMBLY,ref,alt,safe_trans,gene_id,INTERGENIC,DEC_CHILD_SEX,ACCESS,inher_stat,genotype)
out_cntr += 1
out_han.write(to_write)
# write to the IGV file
i_s = pos - SNAP_FLANK
i_e = pos + SNAP_FLANK
# check if above FS/SOR_THRESH to include in the snapshot name
if (FS == '') or (SOR == ''):
flag = 'NA'
elif (FS >= FS_THRESH) and (SOR >= SOR_THRESH):
flag = 'FS_%.1f_SOR_%.1f' % (FS,SOR)
else:
flag = 'OK'
i_name = '%s_%s_%s_%s_%s_%s.png' % (CHILD_ID,chr,pos,ref,alt,flag)
out_igv_han.write('goto %s:%s-%s\n' % (chr,i_s,i_e))
out_igv_han.write('sort strand\n')
out_igv_han.write('squish\n')
out_igv_han.write('snapshot %s\n' % (i_name))
out_igv_han.write('\n')
else:
print "Cannot establish the type of this VCF variant"
print line
raise SystemExit
in_han.close()
out_han.close()
out_igv_han.close()
### check if all G2P and VASE variants were found/matched in the proband's VCF
found_all_G2P = True
for k,v in G2P_DICT.iteritems():
if int(v) == 0:
print k
found_all_G2P = False
break
if found_all_G2P:
print "OK: Found all %s G2P variants in the proband's VCF file" % (len(G2P_DICT))
else:
print "ERROR: Could not find all G2P variants in the probands VCF file"
raise SystemExit
### check if all G2P variants are written out
if out_cntr == NUM_UNIQ_G2P_VARS:
print "OK: All G2P vars are recorded in the output DECIPHER file"
else:
print "ERROR: *NOT* all G2P vars are recorded in the G2P VCF file"
print "Wrote %s variants in outfile = %s" % (out_cntr,out_dec_file)
print "The batch snapshot file = %s" % (out_igv_file)
sys.stdout.flush()
def read_all_VCF_vars(in_vcf_file,THIS_DICT):
in_han = gzip.open(in_vcf_file,'r')
for line in in_han:
if line.startswith('#'):
continue
data = [x.strip() for x in line.strip().split('\t')]
chr = data[0]
pos = int(data[1])
ref = data[3]
alt = data[4]
# did the splitting and normalizing - should not have multiallelic variants
if alt.find(',') != -1:
print "ERROR: found multiallelic variant"
print line
raiseSystemExit
key = '%s:%s:%s:%s' % (chr,pos,ref,alt)
if key not in THIS_DICT:
THIS_DICT[key] = 1
else:
print "ERROR: duplicate key = %s in %s" % (key,in_vcf_file)
raise SystemExit
in_han.close()
def read_G2P(in_file):
global NUM_UNIQ_G2P_VARS
known_OBS_states = ['monoallelic_autosomal','biallelic_autosomal','monoallelic_X_hem','monoallelic_X_het']
# first, read the G2P variants on canonical transcripts for the singleton
CHILD_DICT = defaultdict(dict) # 1st level key: OBS state; 2nd level key: chr:start:end:ref:alt; value: (ZYG,gene,trans)
in_han = open(in_file,'r')
for line in in_han:
data = [x.strip() for x in line.strip().split('\t')]
# get the individual_id
sam_id = data[0]
# ignore variants not on canonical transcripts
is_canon = data[3]
if is_canon != 'is_canonical':
continue
# split the variants based on the gene's OBS model of inheritance
inher_model = data[4]
aaa,OBS_state = inher_model.split('=')
if OBS_state not in known_OBS_states:
print "ERROR: unknown OBS state = %s in %s" % (OBS_state,in_file)
raise SystemExit
# get the gene name in format ENSG00000165899(C12orf64,OTOGL) or gene-MYT1L(MYT1L)
gene_name = data[1]
# get the transcript name in format ENST00000238647 or gene-MYT1L(MYT1L)
transcript = data[2]
# this is a list of variants (n>=1) on a canonical transcript in a gene being considered under any OBS state
var_list = [y.strip() for y in data[6].split(';')]
for v in var_list:
v_details = [z.strip() for z in v.split(':')]
chr = v_details[0]
start = int(v_details[1])
end = int(v_details[2])
ref = v_details[3]
alt = v_details[4]
GT = v_details[5]
second_key = '%s:%s:%s:%s:%s' % (chr,start,end,ref,alt)
if sam_id == CHILD_ID:
# check for duplication
if OBS_state not in CHILD_DICT:
CHILD_DICT[OBS_state][second_key] = (GT,gene_name,transcript)
elif second_key not in CHILD_DICT[OBS_state]:
CHILD_DICT[OBS_state][second_key] = (GT,gene_name,transcript)
else: # already recorded this variant
# if we have refseq recorded and this is ensembl --> replace
if not CHILD_DICT[OBS_state][second_key][1].startswith('ENSG'): # recorded is refseq
if gene_name.startswith('ENSG'): # this is ensembl
CHILD_DICT[OBS_state][second_key] = (GT,gene_name,transcript) # replace
else: # this is refseq again, ignore
pass
else: # recorded is ensembl, ignore
pass
else:
print "ERROR: cannot identify the person for this variant"
print line
raise SystemExit
in_han.close()
### print out the number of unique G2P variants in CHILD ###
child_mono = 0
child_bi = 0
child_hem = 0
child_het = 0
if 'monoallelic_autosomal' in CHILD_DICT:
child_mono = len(CHILD_DICT['monoallelic_autosomal'])
if 'biallelic_autosomal' in CHILD_DICT:
child_bi = len(CHILD_DICT['biallelic_autosomal'])
if 'monoallelic_X_hem' in CHILD_DICT:
child_hem = len(CHILD_DICT['monoallelic_X_hem'])
if 'monoallelic_X_het' in CHILD_DICT:
child_het = len(CHILD_DICT['monoallelic_X_het'])
print "CHILD (%s): number of unique G2P variants on canon transcript in the following OBS states" % (CHILD_ID)
print " monoallelic_autosomal: %s" % (child_mono)
print " biallelic_autosomal: %s" % (child_bi)
print " monoallelic_X_hem: %s" % (child_hem)
print " monoallelic_X_het: %s" % (child_het)
######################################################################################################
#### Dominant filtering ####
#### if the gene has been considered under the dominant model (OBS == monoallelic_autosomal) ####
#### exclude child variants seen in UNAFFECTED mother/father, regardless of GT ####
######################################################################################################
print ""
print "=== monoallelic autosomal (DOMINANT) filtering ==="
for key in CHILD_DICT['monoallelic_autosomal']: # this the second key: chr:start:end:ref:alt; value: (ZYG,gene,trans)
CHILD_GT = CHILD_DICT['monoallelic_autosomal'][key][0]
CHILD_GENE = CHILD_DICT['monoallelic_autosomal'][key][1]
CHILD_TRANS = CHILD_DICT['monoallelic_autosomal'][key][2]
# if (key in MOM_DICT['monoallelic_autosomal']) and (MOM_STAT == "UNAFFECTED"):
# MOM_GT = MOM_DICT['monoallelic_autosomal'][key][0]
# print "***[DOMINANT model]*** Excluded CHILD var %s in gene = %s, CHILD_GT = %s, MOM_GT = %s, MOM_STAT = %s" % (key,CHILD_GENE,CHILD_GT,MOM_GT,MOM_STAT)
# continue
#
# if (key in DAD_DICT['monoallelic_autosomal']) and (DAD_STAT == "UNAFFECTED"):
# DAD_GT = DAD_DICT['monoallelic_autosomal'][key][0]
# print "***[DOMINANT model]*** Excluded CHILD var %s in gene = %s, CHILD_GT = %s, DAD_GT = %s, DAD_STAT = %s" % (key,CHILD_GENE,CHILD_GT,DAD_GT,DAD_STAT)
# continue
# if a non-normalized INDEL in child G2P - must adjust (should not happen really, we split, normalized and left-aligned the family VCF before sending it to VEP+G2P)
chr,start,end,ref,alt = key.split(":")
if len(ref) > 1 and len(alt) > 1: # an INDEL - not normalized
if len(ref) < len(alt): # an INS
orig_start = start
orig_ref = ref
orig_alt = alt
start = orig_start
ref = '-'
alt = orig_alt[len(orig_ref):]
print " WARNING: original INS = %s:%s:%s:%s:%s --> replaced with INS = %s:%s:%s:%s" % (chr,orig_start,end,orig_ref,orig_alt,chr,start,ref,alt)
else: # a DEL
print "ERROR: At the momemnt, cannot deal with this non-normalized deletion"
print line
raise SystemExit
new_key = '%s:%s:%s:%s' % (chr,start,ref,alt)
# record the data for CHILD G2P variants (for OBS=monoallelic)
if new_key not in G2P_DICT:
G2P_DICT[new_key] = 0
else:
# print "ERROR: duplicate G2P variant new_key = %s" % (new_key)
# raise SystemExit
# this will happen if a gene is e.g. hemizygous,x-linked dominant - there will be two separate lines in the output for each req
pass
# and record the required data (CHILD_TRANS,CHILD_GENE,CHILD_GT) in G2P_DATA
if new_key not in G2P_DATA:
G2P_DATA[new_key] = (CHILD_TRANS,CHILD_GENE,CHILD_GT)
else:
# print "ERROR: duplicate G2P variant new_key = %s" % (new_key)
# raise SystemExit
# this will happen if a gene is e.g. hemizygous,x-linked dominant - there will be two separate lines in the output for each req
pass
NUM_UNIQ_G2P_VARS = len(G2P_DICT)
print "Found %s unique G2P variants in CHILD (%s) after considering MONOALLELIC genes" % (NUM_UNIQ_G2P_VARS,CHILD_ID)
sys.stdout.flush()
print ""
##############################################################################################################
#### Recessive filtering ####
#### under the recessive model (OBS == biallelic_autosomal) - consider ALL variants per gene ####
#### must all be HET in CHILD, GT in parent does not matter ####
#### all of them must *clearly* come from only one of the parents (maternally/paternally + biparental) ####
#### and this parent must be unaffected ####
#### if all these: then exclude all child variants in this gene ####
##############################################################################################################
print ""
print "=== biallelic autosomal (RECESSIVE) filtering ==="
GENE_KEY_GT = defaultdict(dict) # for child - 1st level key: gene_name; 2nd level key: chr:start:end:ref:alt; value: (GT,trans)
# process all variants in biallelic genes in child
for key in CHILD_DICT['biallelic_autosomal']: # this the second key: chr:start:end:ref:alt; value: (ZYG,gene,trans)
b_GT = CHILD_DICT['biallelic_autosomal'][key][0]
b_gene = CHILD_DICT['biallelic_autosomal'][key][1]
b_trans = CHILD_DICT['biallelic_autosomal'][key][2]
GENE_KEY_GT[b_gene][key] = (b_GT,b_trans)
# iterate over genes in GENE_KEY_GT
for g in GENE_KEY_GT: # this is the biallelic gene name
# all_HET = True
#
# # iterate over variants in this gene
# for kx in GENE_KEY_GT[g]: # this the second key: chr:start:end:ref:alt
# if GENE_KEY_GT[g][kx][0] == 'HOM': # there is a HOM variant in the child - NO filtering
# all_HET = False
# break
#
# if all_HET: # for this gene
# # all variants in this gene in the CHILD are HET, check if all come from a single unaffected parent
# # if yes, filter out and write a message to the log file
# # if not, to be added to G2P_DICT and G2P_DATA for further processing
#
# all_from_one_parent = True
#
# # iterate again over the variants in this gene
# VAR_SOURCE_LIST = {} # key: chr:start:end:ref:alt in child; value: (NONE) or (MOM or DAD or BOTH and the parent is UNAFFECTED)
#
# for ky in GENE_KEY_GT[g]: # this the second key: chr:start:end:ref:alt
#
# this_var_status = 'NONE'
#
# if ((ky in MOM_DICT['biallelic_autosomal']) or (ky in MOM_DICT['monoallelic_autosomal'])) and (MOM_STAT == "UNAFFECTED"):
# this_var_status = 'MOM'
# if ((ky in DAD_DICT['biallelic_autosomal']) or (ky in DAD_DICT['monoallelic_autosomal'])) and (DAD_STAT == "UNAFFECTED"):
# if this_var_status == 'NONE':
# this_var_status = 'DAD'
# elif this_var_status == 'MOM':
# this_var_status = 'BOTH'
#
# VAR_SOURCE_LIST[ky] = this_var_status
#
# # have collected the parent source for all variants in this gene
# tot_num_vars = len(VAR_SOURCE_LIST)
# num_mom = 0
# num_dad = 0
# num_none = 0
# for kt,v in VAR_SOURCE_LIST.iteritems():
# if v == 'NONE':
# num_none += 1
# elif v == 'MOM':
# num_mom += 1
# elif v == 'DAD':
# num_dad += 1
# elif v == 'BOTH':
# num_mom += 1
# num_dad += 1
# else:
# print "ERROR: cannot understand the source parent = %s" % (v)
# raise SystemExit
#
# if num_none > 0:
# all_from_one_parent = False
# elif num_mom < tot_num_vars and num_dad < tot_num_vars:
# all_from_one_parent = False
#
# # if all variants in the child in this gene are found in single unaffected parent - filter out
# if all_from_one_parent:
# for kz in GENE_KEY_GT[g]:
# print "***[RECESSIVE model]*** Excluded CHILD HET var %s in gene = %s, found in = %s, PARENT_STAT = UNAFFECTED" % (kz,g,VAR_SOURCE_LIST[kz])
# continue
#
# # end processing all HET variants in the proband - if all from single unaffected parent they have been excluded, message to the log written
# # and gone to evaluating the next biallelic gene in the child
#
# # if here
# # - either not all CHILD variants in this gene are not HET, or
# # - not all of them can be traced to a single unaffected parent
# # --> add to be processed
#
# here we are at gene level, must iterate over all variants in this gene
# iterate over variants in this gene
for kkk in GENE_KEY_GT[g]: # this the second key: chr:start:end:ref:alt
CHILD_GT = CHILD_DICT['biallelic_autosomal'][kkk][0]
CHILD_GENE = CHILD_DICT['biallelic_autosomal'][kkk][1]
CHILD_TRANS = CHILD_DICT['biallelic_autosomal'][kkk][2]
# if a non-normalized INDEL in child G2P - must adjust (should not happen really, we split, normalized and left-aligned the family VCF before sending it to VEP+G2P)
chr,start,end,ref,alt = kkk.split(":")
if len(ref) > 1 and len(alt) > 1: # an INDEL - not normalized
if len(ref) < len(alt): # an INS
orig_start = start
orig_ref = ref
orig_alt = alt
start = orig_start
ref = '-'
alt = orig_alt[len(orig_ref):]
print " WARNING: original INS = %s:%s:%s:%s:%s --> replaced with INS = %s:%s:%s:%s" % (chr,orig_start,end,orig_ref,orig_alt,chr,start,ref,alt)
else: # a DEL
print "ERROR: At the momemnt, cannot deal with this non-normalized deletion"
print line
raise SystemExit
new_key = '%s:%s:%s:%s' % (chr,start,ref,alt)
# record the data for CHILD G2P variants (for OBS=biallelic)
if new_key not in G2P_DICT:
G2P_DICT[new_key] = 0
else:
# print "ERROR: duplicate G2P variant new_key = %s" % (new_key)
# raise SystemExit
# this will happen if a gene is e.g. hemizygous,x-linked dominant - there will be two separate lines in the output for each req
pass
# and record the required data (CHILD_TRANS,CHILD_GENE,CHILD_GT) in G2P_DATA
if new_key not in G2P_DATA:
G2P_DATA[new_key] = (CHILD_TRANS,CHILD_GENE,CHILD_GT)
else:
# print "ERROR: duplicate G2P variant new_key = %s" % (new_key)
# raise SystemExit
# this will happen if a gene is e.g. hemizygous,x-linked dominant - there will be two separate lines in the output for each req
pass
NUM_UNIQ_G2P_VARS = len(G2P_DICT)
print "Found %s unique G2P variants in CHILD (%s) after considering MONOALLELIC and BIALLELIC genes" % (NUM_UNIQ_G2P_VARS,CHILD_ID)
sys.stdout.flush()
print ""
####################################################################################################################
#### X-linked filtering ####
#.# #### under the x-linked model (OBS == hemizygous or x-linked dominant, but NOT x-linked over-dominance) ####
#### under the chrX model (OBS == monoallelic_X_hem or monoallelic_X_het) ####
#### exclude child HET variants if seen as HOM in UNAFFECTED father ####
#### ####
#### Note 18/01/2022 ####
#### This is a temporary solution, since x-linked dominant and x-linked over-dominance -> monoallelic_X_het ####
#### and we should filter x-linked dominant and monoallelic_X_hem, but not x-linked over-dominance ####
#### the code below treats x-linked over-dominance as the others (i.e. filters, while it should not) ####
#### Issue flagged to G2P plug-in team, awaiting their fix ####
#### for now manually scan the output of G2P for the proband (both for boys and girls) ####
#### to check if any variant has been called in PCDH19 and EFNB1 ####
#### also for all the variants filtered out from monoallelic_X_het we will print in the log the gene name ####
####################################################################################################################
print ""
print "=== X-linked filtering ==="
#######################################
### process monoallelic_X_hem genes ###
#######################################
for key in CHILD_DICT['monoallelic_X_hem']: # this the second key: chr:start:end:ref:alt; value: (ZYG,gene,trans)
CHILD_GT = CHILD_DICT['monoallelic_X_hem'][key][0]
CHILD_GENE = CHILD_DICT['monoallelic_X_hem'][key][1]
CHILD_TRANS = CHILD_DICT['monoallelic_X_hem'][key][2]
# if CHILD_GT == 'HOM': # do NOT filter HOM variants in proband (i.e., hemizygous in boy or HOM in girl)
# pass
# else:
# if (key in DAD_DICT['monoallelic_X_hem']) and (DAD_STAT == "UNAFFECTED"):
# DAD_GT = DAD_DICT['monoallelic_X_hem'][key][0]
# if DAD_GT == 'HOM': # i.e., hemizygous variant in unaffected father
# print "***[monoallelic_X_hem]*** Excluded CHILD var %s in gene = %s, CHILD_GT = %s, DAD_GT = %s, DAD_STAT = %s" % (key,CHILD_GENE,CHILD_GT,DAD_GT,DAD_STAT)
# continue
#
# if a non-normalized INDEL in child G2P - must adjust (should not happen really, we split, normalized and left-aligned the family VCF before sending it to VEP+G2P)
chr,start,end,ref,alt = key.split(":")
if len(ref) > 1 and len(alt) > 1: # an INDEL - not normalized
if len(ref) < len(alt): # an INS
orig_start = start
orig_ref = ref
orig_alt = alt
start = orig_start
ref = '-'
alt = orig_alt[len(orig_ref):]
print " WARNING: original INS = %s:%s:%s:%s:%s --> replaced with INS = %s:%s:%s:%s" % (chr,orig_start,end,orig_ref,orig_alt,chr,start,ref,alt)
else: # a DEL
print "ERROR: At the momemnt, cannot deal with this non-normalized deletion"
print line
raise SystemExit
new_key = '%s:%s:%s:%s' % (chr,start,ref,alt)
# record the data for CHILD G2P variants (for OBS=monoallelic_X_hem)
if new_key not in G2P_DICT:
G2P_DICT[new_key] = 0
else:
# print "ERROR: duplicate G2P variant new_key = %s" % (new_key)
# raise SystemExit
# this will happen if a gene is e.g. hemizygous,x-linked dominant - there will be two separate lines in the output for each req
pass
# and record the required data (CHILD_TRANS,CHILD_GENE,CHILD_GT) in G2P_DATA
if new_key not in G2P_DATA:
G2P_DATA[new_key] = (CHILD_TRANS,CHILD_GENE,CHILD_GT)
else:
# print "ERROR: duplicate G2P variant new_key = %s" % (new_key)
# raise SystemExit
# this will happen if a gene is e.g. hemizygous,x-linked dominant - there will be two separate lines in the output for each req
pass
#######################################
### process monoallelic_X_het genes ###
#######################################
for key in CHILD_DICT['monoallelic_X_het']: # this the second key: chr:start:end:ref:alt; value: (ZYG,gene,trans)
CHILD_GT = CHILD_DICT['monoallelic_X_het'][key][0]
CHILD_GENE = CHILD_DICT['monoallelic_X_het'][key][1]
CHILD_TRANS = CHILD_DICT['monoallelic_X_het'][key][2]
# if CHILD_GT == 'HOM': # do NOT filter HOM variants (i.e., hemizygous in boy or HOM in girl)
# pass
# else:
# if (key in DAD_DICT['monoallelic_X_het']) and (DAD_STAT == "UNAFFECTED"):
# DAD_GT = DAD_DICT['monoallelic_X_het'][key][0]
# if DAD_GT == 'HOM': # i.e., x-linked dominant variant in unnafected father
# print "***[monoallelic_X_het]*** Excluded CHILD var %s in gene = %s, CHILD_GT = %s, DAD_GT = %s, DAD_STAT = %s" % (key,CHILD_GENE,CHILD_GT,DAD_GT,DAD_STAT)
# continue
#
# if a non-normalized INDEL in child G2P - must adjust (should not happen really, we split, normalized and left-aligned the family VCF before sending it to VEP+G2P)
chr,start,end,ref,alt = key.split(":")
if len(ref) > 1 and len(alt) > 1: # an INDEL - not normalized
if len(ref) < len(alt): # an INS
orig_start = start
orig_ref = ref
orig_alt = alt
start = orig_start
ref = '-'
alt = orig_alt[len(orig_ref):]
print " WARNING: original INS = %s:%s:%s:%s:%s --> replaced with INS = %s:%s:%s:%s" % (chr,orig_start,end,orig_ref,orig_alt,chr,start,ref,alt)
else: # a DEL
print "ERROR: At the momemnt, cannot deal with this non-normalized deletion"
print line
raise SystemExit
new_key = '%s:%s:%s:%s' % (chr,start,ref,alt)
# record the data for CHILD G2P variants (for OBS=monoallelic_X_het)
if new_key not in G2P_DICT:
G2P_DICT[new_key] = 0
else:
# print "ERROR: duplicate G2P variant new_key = %s" % (new_key)
# raise SystemExit
# this will happen if a gene is e.g. hemizygous,x-linked dominant - there will be two separate lines in the output for each req
pass
# and record the required data (CHILD_TRANS,CHILD_GENE,CHILD_GT) in G2P_DATA
if new_key not in G2P_DATA:
G2P_DATA[new_key] = (CHILD_TRANS,CHILD_GENE,CHILD_GT)
else:
# print "ERROR: duplicate G2P variant new_key = %s" % (new_key)
# raise SystemExit
# this will happen if a gene is e.g. hemizygous,x-linked dominant - there will be two separate lines in the output for each req
pass
##.# ########################################################################
##.# ### process x-linked over-dominance genes - no filtering to be done ###
##.# ########################################################################
#
##.# for key in CHILD_DICT['x-linked over-dominance']: # this the second key: chr:start:end:ref:alt; value: (ZYG,gene,trans)
#
##.# CHILD_GT = CHILD_DICT['x-linked over-dominance'][key][0]
##.# CHILD_GENE = CHILD_DICT['x-linked over-dominance'][key][1]
##.# CHILD_TRANS = CHILD_DICT['x-linked over-dominance'][key][2]
#
##.# # if a non-normalized INDEL in child G2P - must adjust (should not happen really, we split, normalized and left-aligned the family VCF before sending it to VEP+G2P)
##.# chr,start,end,ref,alt = key.split(":")
##.# if len(ref) > 1 and len(alt) > 1: # an INDEL - not normalized
##.# if len(ref) < len(alt): # an INS
##.# orig_start = start
##.# orig_ref = ref
##.# orig_alt = alt
##.# start = orig_start
##.# ref = '-'
##.# alt = orig_alt[len(orig_ref):]
##.# print " WARNING: original INS = %s:%s:%s:%s:%s --> replaced with INS = %s:%s:%s:%s" % (chr,orig_start,end,orig_ref,orig_alt,chr,start,ref,alt)
##.# else: # a DEL
##.# print "ERROR: At the momemnt, cannot deal with this non-normalized deletion"
##.# print line
##.# raise SystemExit
#
##.# new_key = '%s:%s:%s:%s' % (chr,start,ref,alt)
#
##.# # record the data for CHILD G2P variants (for OBS=x-linked over-dominance)