-
Notifications
You must be signed in to change notification settings - Fork 0
/
methyl_seq_bak.R
1570 lines (1281 loc) · 80.8 KB
/
methyl_seq_bak.R
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
library(data.table)
library(log4r)
library(RSQLite)
library(DBI)
library(ggplot2)
library(RSQLite)
options(width=155)
options(scipen=999) ## dont write number as scientific note
cfg = new.env()
cfg$assay = 'methylation'
cfg$species = 'GRCh37'
cwd = '/ifs/work/solitlab/huw/solit/study/hiseq/Project_06000_DD'
cfg$cwd = cwd
cfg$resDir = paste0(cwd, '/r_fang')
cfg$pre = 'scc'
cfg$assay = 'methylation'
delFiles = T
source('~/pipeline/configure_env.R')
setwd(cwd)
cwd
fread('target.tsv') -> target
target[, R12 := 'R1']
target[grep("R2_001", fastq), R12 := 'R2']
target[, samplename := sub(".*\\/(.*)_IGO.*", "\\1", fastq)]
target[, samplename := sub("_test", "", samplename)]
target[, samplename := paste0("pat-", samplename)]
target
target = dcast(target, samplename ~ R12, value.var = 'fastq')
target[, patientname := substr(samplename, 1, 7)]
target
target[, sampledir := paste0(cfg$initDir, '/', samplename)]
target[, {system(paste('mkdir -p ', sampledir))}, by=1:nrow(target)]
dir.exists(target$sampledir)
## fastqc
fastqc.jobname = paste0("fastqc")
fastq = paste(c(target$R1, target$R2), collapse=" ")
fastqc.cmd = bsub.head(fastqc.jobname, mem=10, cpu=20)
fastqc.cmd = paste0(fastqc.cmd, " \"", cfg$fastqc, " -o ", cfg$fastqcDir, " -j ", cfg$java, " -f fastq -t 20 ", fastq, " \"")
fastqc.cmd
exe.jobs(fastqc.cmd, logger)
## make genome for bismark
bismark.db.jobname = 'bismark.db'
bismark.db.cmd = bsub.head(bismark.db.jobname, mem=20, cpu=10)
bismark.db.cmd = paste0(bismark.db.cmd, " \"", cfg$bismark, '/bismark_genome_preparation --path_to_bowtie ', cfg$bowtie2dir, ' --verbose /ifs/work/solitlab/huw/study/db/hsa/grch37_bismark \"')
bismark.db.cmd
exe.jobs(bismark.db.cmd, cfg$logger)
## align
target[, bismark.jobname := paste0(samplename, '.bismark')]
target[, bamfile := paste0(cfg$sampledir, '/', cfg$samplename, '_pe.bam')]
target[, bismark.cmd := bsub.head(bismark.jobname, mem=60, cpu=30, postdone = 'bismark.db', We = '30:00')]
target[, bismark.cmd := paste0(bismark.cmd, " \"", cfg$bismark, "/bismark --path_to_bowtie ", cfg$bowtie2dir, " --quiet -o ", sampledir)]
target[, bismark.cmd := paste0(bismark.cmd, " --samtools_path ", cfg$samtoolsdir, " -B ", samplename, " --nucleotide_coverage ")]
target[, bismark.cmd := paste0(bismark.cmd, " ", cfg$bismarkGRCh37, " -1 ", R1, " -2 ", R2, " \"")]
target[1, bismark.cmd]
target$bamfile
colnames(target)
exe.jobs(target$bismark.cmd, cfg$logger)
## deduplicate
#target[, sorted.bam.file := paste0(sampledir, '/', samplename, '_pe_sorted.bam')]
target[, bamfile := paste0(sampledir, '/', samplename, '_pe.bam')]
target[, dedup.bam.file := paste0(sampledir, '/', samplename, '.deduplicated_pe.bam')]
target[, dedup.jobname := paste0(samplename, '.dedup')]
target[, dedup.cmd := bsub.head(dedup.jobname, mem=30, cpu=10, postdone = bismark.jobname, We = '30:00'), by=1:nrow(target)]
target[, dedup.cmd := paste0(dedup.cmd, " \"", cfg$bismark, '/deduplicate_bismark -p --output_dir ', sampledir, ' --bam --samtools_path ', cfg$samtools, ' ', bamfile, "\"")]
target[2, dedup.cmd]
exe.jobs(target$dedup.cmd, cfg$logger)
## sort bam file
target[, sorted.bam.file := paste0(sampledir, '/', samplename, '_pe.deduplicated_sorted.bam')]
target[, bam.sort.jobname := paste0(samplename, '.sort.bam')]
target[, bam.sort.cmd := bsub.head(bam.sort.jobname, mem=60, cpu=30, postdone = dedup.jobname, We = '30:00'), by=1:nrow(target)]
target[, bam.sort.cmd := paste0(bam.sort.cmd, " \"", cfg$JAVA, '/java -Xms256M -Xmx60g -Djava.io.tmpdir=', cfg$tmpdir, ' -jar ', cfg$PICARD, ' SortSam I=', dedup.bam.file, " O=", sorted.bam.file, " CREATE_INDEX=true SORT_ORDER=coordinate\"")]
target[2, bam.sort.cmd]
exe.jobs(target$bam.sort.cmd, cfg$logger)
## collect alignment summary metrics
target[, align.sum.file := paste0(sampledir, '/', samplename, '_align_sum.txt')]
target[, align.sum.jobname := paste0(samplename, '.align.sum')]
target[, align.sum.cmd := bsub.head(align.sum.jobname, mem=10, cpu=3, postdone = bam.sort.jobname, We = '10:00'), by = 1:nrow(target)]
target[, align.sum.cmd := paste0(align.sum.cmd, " \"", cfg$JAVA, '/java -Xms256M -Xmx10g -Djava.io.tmpdir=', cfg$tmpdir, ' -jar ', cfg$PICARD, ' CollectAlignmentSummaryMetrics R=', cfg$genomeGRCh37Fasta, " I=", sorted.bam.file, " O=", align.sum.file, " \"")]
target[2, align.sum.cmd]
exe.jobs(target$align.sum.cmd, cfg$logger)
## collect gc bias metrics
target[, gc.bias.metrics.file := paste0(sampledir, '/', samplename, '_gc_bias_metrics.txt')]
target[, gc.bias.metrics.pdf.file := paste0(sampledir, '/', samplename, '_gc_bias_metrics.pdf')]
target[, gc.bias.metrics.sum.file := paste0(sampledir, '/', samplename, '_gc_bias_metrics_sum.txt')]
target[, gc.bias.jobname := paste0(samplename, '.gc.bias')]
target[, gc.bias.cmd := bsub.head(gc.bias.jobname, mem=10, cpu=3, postdone = bam.sort.jobname, We = '10:00'), by = 1:nrow(target)]
target[, gc.bias.cmd := paste0(gc.bias.cmd, " \"", cfg$JAVA, '/java -Xms256M -Xmx10g -Djava.io.tmpdir=', cfg$tmpdir, ' -jar ', cfg$PICARD, ' CollectGcBiasMetrics R=', cfg$genomeGRCh37Fasta, " I=", sorted.bam.file, " O=", gc.bias.metrics.file, " CHART=", gc.bias.metrics.pdf.file, " S=", gc.bias.metrics.sum.file, "\"")]
target[1, gc.bias.cmd]
exe.jobs(target$gc.bias.cmd, cfg$logger)
## collect hs metrics
# mmuMethylInterval = '/home/huw/program/truseq-methyl-capture-epic-manifest-file.bed'
target[, hs.metrics.file := paste0(sampledir, '/', samplename, '_hs_metrics.txt')]
target[, hs.jobname := paste0(samplename, '.hs')]
target[, hs.cmd := bsub.head(hs.jobname, mem=10, cpu=3, postdone = bam.sort.jobname, We = '10:00'), by = 1:nrow(target)]
target[, hs.cmd := paste0(hs.cmd, " \"", cfg$JAVA, '/java -Xms256M -Xmx10g -Djava.io.tmpdir=', cfg$tmpdir, ' -jar ', cfg$PICARD, ' CollectHsMetrics R=', cfg$genomeGRCh37Fasta, " I=", sorted.bam.file, " O=", hs.metrics.file, " TARGET_INTERVALS=", cfg$hsaMethylTargetInterval, " BAIT_INTERVALS=", cfg$hsaMethylBaitInterval, "\"")]
target[1, hs.cmd]
file.exists(cfg$hsaMethylTargetInterval)
exe.jobs(target$hs.cmd, cfg$logger)
## collect insert size
target[, insert.size.metrics.file := paste0(sampledir, '/', samplename, '_insert_size_metrics.txt')]
target[, insert.size.metrics.histgram.file := paste0(sampledir, '/', samplename, '_insert_size_metrics_histgram.pdf')]
target[, insert.size.jobname := paste0(samplename, '.insert.size')]
target[, insert.size.cmd := bsub.head(insert.size.jobname, mem=10, cpu=3, postdone = bam.sort.jobname, We = '10:00'), by = 1:nrow(target)]
target[, insert.size.cmd := paste0(insert.size.cmd, " \"", cfg$JAVA, '/java -Xms256M -Xmx10g -Djava.io.tmpdir=', cfg$tmpdir, ' -jar ', cfg$PICARD, ' CollectInsertSizeMetrics R=', cfg$genomeGRCh37Fasta, " I=", sorted.bam.file, " O=", insert.size.metrics.file, " H=", insert.size.metrics.histgram.file, " M=0.5\"")]
target[1, insert.size.cmd]
exe.jobs(target$insert.size.cmd, cfg$logger)
## collect targeted pcr metrics
target[, targeted.pcr.metrics.file := paste0(sampledir, '/', samplename, '_targeted.pcr_metrics.txt')]
target[, targeted.pcr.metrics.histgram.file := paste0(sampledir, '/', samplename, '_targeted.pcr_metrics_histgram.pdf')]
target[, targeted.pcr.jobname := paste0(samplename, '.targeted.pcr')]
target[, targeted.pcr.cmd := bsub.head(targeted.pcr.jobname, mem=10, cpu=3, postdone = bam.sort.jobname, We = '10:00'), by = 1:nrow(target)]
target[, targeted.pcr.cmd := paste0(targeted.pcr.cmd, " \"", cfg$JAVA, '/java -Xms256M -Xmx10g -Djava.io.tmpdir=', cfg$tmpdir, ' -jar ', cfg$PICARD, ' CollectInsertSizeMetrics R=', cfg$genomeGRCh37Fasta, " I=", sorted.bam.file, " O=", targeted.pcr.metrics.file, " H=", targeted.pcr.metrics.histgram.file, " M=0.5\"")]
target[1, targeted.pcr.cmd]
exe.jobs(target$targeted.pcr.cmd, cfg$logger)
## bismark_methylation_extractor
target[, bismark.ex.jobname := paste0(samplename, ".bismark.ex")]
target[, bismark.ex.cmd := bsub.head(bismark.ex.jobname, mem=30, We ='10:00', cpu=22, postdone = dedup.jobname), by=1:nrow(target)]
target[, bismark.ex.cmd := paste0(bismark.ex.cmd, " \"", cfg$bismark, "/bismark_methylation_extractor --samtools_path ", cfg$samtoolsdir, " -o ", sampledir, " --parallel 5 --bedGraph")]
target[, bismark.ex.cmd := paste0(bismark.ex.cmd, " ", dedup.bam.file, " \"")]
target[1, bismark.ex.cmd]
exe.jobs(target$bismark.ex.cmd, cfg$logger)
## bismark2report
#target[, report.file := paste0(sampledir, '/', samplename, '_PE_report.txt')]
target[, dedup.report.file := paste0(sampledir, '/', samplename, '_pe.deduplicated_report.txt')]
target[, bismark.report.jobname := paste0(samplename, ".bismark.report")]
target[, bismark.report.cmd := bsub.head(bismark.report.jobname, mem=2, We ='10:00', cpu=1, postdone = bismark.jobname), by=1:nrow(target)]
target[, bismark.report.cmd := paste0(bismark.report.cmd, " \"", cfg$bismark, "/bismark2report --dir ", sampledir, " --alignment_report ", dedup.report.file, "\"")]
target[1, bismark.report.cmd]
exe.jobs(target$bismark.report.cmd, cfg$logger)
## not run
## bismark2summary
target[, summary.file := paste0(sampledir, '/', samplename, '_PE_summary.txt')]
target[, bismark.summary.jobname := paste0(samplename, ".bismark.summary")]
target[, bismark.summary.cmd := bsub.head(bismark.summary.jobname, mem=5, We ='10:00', cpu=1, postdone = ), by=1:nrow(target)]
target[, bismark.summary.cmd := paste0(bismark.summary.cmd, " \"", bismark, "/bismark2summary -o ", sampledir, "/", samplename, "_pe ", dedup.bam.file, "\"")]
target[1, bismark.summary.cmd]
target$dedup.bam.file
exe.jobs(target$bismark.summary.cmd, cfg$logger)
target$dedup.bam.file
## methylkit
library(methylKit)
target.interval = fread(cfg$hsaMethylBaitIntervalOrig, autostart=48)
setkey(target.interval, V1, V2, V3)
save.image()
## convert the bismark results for methylkit
## min coverage > 10
## methylkit.file for methylkit library
## methylkit.bed.file for calling C/G for strand information
## methylkit.base.file the actual C/G base data
## revise the methylkit.file to add the strand information
target[, cov.file := paste0(sampledir, '/', samplename, '_pe.deduplicated.bismark.cov.gz')]
target[, methylkit.file := paste0(sampledir, '/', samplename, '_pe.deduplicated.bismark_methylkit.mincov10.txt')]
file.exists(target$cov.file)
target[, {
met.cov = fread(paste0("zcat ", cov.file))
met.cov = met.cov[(V5 + V6) > 10,]
met.cov[, V1 := paste0('chr', V1)]
met.cov = foverlaps(met.cov, target.interval[, 1:3])
head(met.cov)
met.cov = met.cov[!is.na(V2),]
met.cov[, coverage := V5 + V6]
met.cov[, freqT := round(100 * V5 / coverage, 2)]
met.cov[, freqC := round(100 * V6 / coverage, 2)]
met.cov[, chrBase := paste0(V1, '.', i.V2)]
met.cov[, strand := 'F']
setnames(met.cov, 'i.V2', 'base')
setnames(met.cov, 'V1', 'chr')
met.cov = met.cov[, .(chrBase, chr, base, strand, coverage, freqC, freqT)]
fwrite(met.cov, file=methylkit.file, sep="\t", quote=F)
}, by=1:nrow(target)]
## decide the strand for bismark coverage results
target$samplename
target[, methylkit.bed.file := paste0(sampledir, '/', samplename, '_pe.deduplicated.bismark_methylkit.mincov10.bed')]
target[, {
cmd = paste0("awk '{if(NR==1){}else{print($2, $3-1, $3)}}' ", methylkit.file, " | sed \"s/ /\t/g\" | sort -k1 -k2 | sed \"s/chr//\" > ", methylkit.bed.file)
cmd
cat(samplename, " now ... \n\t", cmd)
system(cmd)
}, by=1:nrow(target)]
target[, getCG.jobname := paste0(samplename, '.getCG')]
target[, base.gc.file := paste0(sampledir, '/', samplename, '_base_GC.txt')]
target[, getCG.cmd := bsub.head(getCG.jobname, mem=20, cpu=1, We='10:00')]
target[, getCG.cmd := paste0(getCG.cmd, " \"", cfg$bedtools, '/bedtools getfasta -fi ', cfg$genomeGRCh37Fasta, ' -bed ', methylkit.bed.file, ' -fo ', base.gc.file, ' -tab ', "\"")]
target[1, getCG.cmd]
exe.jobs(target$getCG.cmd, logger)
## revise the methylkit.file to add the strand information
target$samplename
target[, {
#methylkit.file = target$methylkit.file[3]
#base.gc.file = target$base.gc.file[3]
#rm(methylkit.file, base.gc.file)
cat("\n", samplename, ' now ...\n')
met.cov = fread(methylkit.file, colClass=c(strand='character'))
base.gc = fread(base.gc.file, header=F)
base.gc[, chr := paste0('chr', sub(":.*", "", V1))]
base.gc[, base := sub(".*-", "", V1)]
base.gc[, tag := paste0(chr, '.', base)]
merge(met.cov, base.gc[, .(tag, V2)], by.x = 'chrBase', by.y='tag') -> met.cov
met.cov[, strand := 'X']
met.cov[V2 == 'C', strand := 'F']
met.cov[V2 == 'G', strand := 'R']
head(met.cov)
cat('\t total: ', nrow(met.cov), "\n")
cat('\t strand F: ', nrow(met.cov[strand == 'F',]), "\n")
cat('\t strand R: ', nrow(met.cov[strand == 'R',]), "\n")
cat('\t strand X: ', nrow(met.cov[strand == 'X',]), "\n")
if(nrow(met.cov[strand == 'X',]) > 0){
cat(samplename, "\n\tsomething wrong here, row number ", which(met.cov$strand == 'X'))
}
met.cov[, V2 := NULL]
fwrite(met.cov, file= methylkit.file, quote=F, sep="\t")
}, by = 1:nrow(target)]
## not run, segfault, memory not mapped
## methylkit::processBismarkAln.R
library(methylKit)
target[, group := substr(samplename, 9, 10)]
save(target, file='target.RData')
save(cfg, file='cfg.RData')
try{
methyl.o = processBismarkAln(location = as.list(target$sorted.bam.file), sample.id=as.list(target$samplename),
assembly='GRCh37', save.folder=cfg$methylDir, save.context=c("CpG"),
read.context=c("CpG"), nolap = F, mincov=10, minqual=20, phred64=F,
treatment=target$group, save.db=T)
}
try(
{methyl.o = processBismarkAln(location = target$sorted.bam.file[1], sample.id=target$samplename[1],
assembly='GRCh37', save.folder=cfg$methylDir, save.context=c("CpG"),
read.context=c("CpG"), nolap = F, mincov=10, minqual=20, phred64=F) }
)
## use the methylkit.file
save.image()
save(target, cfg, file='apr15.RData')
library(methylKit)
target$samplename
methyl.o = methRead(as.list(target$methylkit.file), sample.id=as.list(target$samplename), assembly='GRCh37', treatment=factor(target$group), context='CpG')
## filter
methyl.fil.o = filterByCoverage(methyl.o, lo.count=10, lo.perc=NULL, hi.count=NULL, hi.perc=100)
rm(methyl.o)
save(methyl.fil.o, file=paste0(cfg$methylDir, '/methyl_filtered_obj.RData'))
## import this raw data into the raw sqlite
library(RSQLite)
library(DBI)
con = dbConnect(drv=SQLite(), dbname=paste0(cfg$methylDir, './sqlite/methyl.raw.sqlite'))
target[, {dt.m = fread(methylkit.file)
dt.m[, samplename := samplename]
dbWriteTable(con, 'methylRaw', as.data.frame(dt.m), append=T)
}, by=1:nrow(target)]
dbSendStatement(con, 'create index iindex on methylRaw (samplename, chr, base)')
dbDisconnet(con)
for(i in 1:nrow(target)){
pdf(file=paste0(cfg$methylDir, '/', target$samplename[i], '.pdf'), width=12, height=6)
getMethylationStats (methyl.fil.o[[i]], plot=T, both.strands=T)
getCoverageStats (methyl.fil.o[[i]], plot=TRUE, both.strands=T)
dev.off()
}
## combine, destrand for both strand when analyzing CpG
meth = unite(methyl.fil.o, destrand=T, min.per.group=1L)
fwrite(meth, file=paste0(cfg$methylDir, '/methylation_matrix.tsv'), sep="\t", quote=F)
save(meth, file=paste0(cfg$methylDir, '/meth_united.RData'))
pdf(file=paste0(cfg$methylDir, '/corr_pca_cluster.pdf'))
getCorrelation(meth, plot=TRUE)
clusterSamples(meth, dist="correlation", method="ward", plot=TRUE)
PCASamples(meth, screeplot=TRUE)
PCASamples(meth)
dev.off()
myDiff = calculateDiffMeth(meth, mc.cores=5)
myDiff0p = getMethylDiff(myDiff, difference=0, qvalue=0.01)
pdf(file=paste0(cfg$methylDir, '/diff_methyl.pdf'))
diffMethPerChr(myDiff, plot=T, qvalue.cutoff=0.01, meth.cutoff=25)
dev.off()
library(genomation)
tx.file = system.file('extdata', 'hg19_gencode_genesymbol.bed.txt', package = 'methylKit')
gene.obj=readTranscriptFeatures(tx.file, up.flank=2000, down.flank=2000)
diffAnn=annotateWithGeneParts(as(myDiff0p, "GRanges"), gene.obj)
diffAnn
head(getAssociationWithTSS(diffAnn))
getTargetAnnotationStats(diffAnn, percentage=TRUE, precedence=TRUE)
cpg.file = system.file('extdata', 'hg19_cpg_island_tag_genesymbol.bed.txt', package = 'methylKit')
cpg.obj = readFeatureFlank(cpg.file, feature.flank.name=c("CpGi","shores"))
# convert methylDiff object to GRanges and annotate
diffCpGann = annotateWithFeatureFlank(as(myDiff0p,"GRanges"), cpg.obj$CpGi,cpg.obj$shores, feature.name = "CpGi",flank.name = "shores")
diffCpGann
pdf(file=paste0(cfg$methylDir, '/dmr_annotation.pdf'))
plotTargetAnnotation(diffAnn,precedence=TRUE, main="DMR")
plotTargetAnnotation(diffCpGann,col=c("green","gray","white"), main="DMR by CpG island")
dev.off()
getFeatsWithTargetsStats(diffAnn,percentage=TRUE)
getFeatsWithTargetsStats(diffCpGann,percentage=TRUE)
## regional DMR promoter
gene.obj$promoters
promoters = regionCounts(methyl.fil.o, gene.obj$promoters)
promoters
promoter.meth = unite(promoters, destrand=T, min.per.group=1L)
promoter.meth
fwrite(promoter.meth, file=paste0(cfg$methylDir, '/methylation_promoter_region_matrix.tsv'), sep="\t", quote=F)
save(promoter.meth, file=paste0(cfg$methylDir, '/promoter_region_meth_united.RData'))
pdf(file=paste0(cfg$methylDir, '/promoter_region_corr_pca_cluster.pdf'))
getCorrelation(promoter.meth, plot=TRUE)
clusterSamples(promoter.meth, dist="correlation", method="ward", plot=TRUE)
PCASamples(promoter.meth, screeplot=TRUE)
PCASamples(promoter.meth)
dev.off()
myDiff.promoter.region = calculateDiffMeth(promoter.meth, mc.cores=3)
myDiff0p.promoter = getMethylDiff(myDiff.promoter.region, difference=0, qvalue=0.01)
rm(myDiff25p.promoter.hi,myDiff25p.promoter.lo)
## regional DMR CpG island
cpgi.region = regionCounts(methyl.fil.o, cpg.obj$CpGi)
cpgi.shore.region = regionCounts(methyl.fil.o, cpg.obj$shores)
cpgi.region.unite = unite(cpgi.region, destrand=T, min.per.group=1L)
cpgi.shore.unite = unite(cpgi.shore.region, destrand=T, min.per.group=1L)
fwrite(cpgi.region.unite, file=paste0(cfg$methylDir, '/methylation_cpgi_region_unite_matrix.tsv'), sep="\t", quote=F)
fwrite(cpgi.shore.unite, file=paste0(cfg$methylDir, '/methylation_cpgi_shore_unite_matrix.tsv'), sep="\t", quote=F)
save(cpgi.region.unite, file=paste0(cfg$methylDir, '/methylation_cpgi_region_matrix.RData'))
save(cpgi.shore.unite, file=paste0(cfg$methylDir, '/methylation_cpgi_shore_matrix.RData'))
pdf(file=paste0(cfg$methylDir, '/cpgi_region_corr_pca_cluster.pdf'))
getCorrelation(cpgi.region.unite, plot=TRUE)
clusterSamples(cpgi.region.unite, dist="correlation", method="ward", plot=TRUE)
PCASamples(cpgi.region.unite, screeplot=TRUE)
PCASamples(cpgi.region.unite)
dev.off()
pdf(file=paste0(cfg$methylDir, '/cpgi_shore_region_corr_pca_cluster.pdf'))
getCorrelation(cpgi.shore.unite, plot=TRUE)
clusterSamples(cpgi.shore.unite, dist="correlation", method="ward", plot=TRUE)
PCASamples(cpgi.shore.unite, screeplot=TRUE)
PCASamples(cpgi.shore.unite)
dev.off()
myDiff.cpgi.region = calculateDiffMeth(cpgi.region.unite, mc.cores=5)
myDiff.cpgi.shore = calculateDiffMeth(cpgi.shore.unite, mc.cores=5)
myDiff0p.cpgi.region = getMethylDiff(myDiff.cpgi.region, difference=0, qvalue=0.01)
myDiff0p.cpgi.shore = getMethylDiff(myDiff.cpgi.shore, difference=0, qvalue=0.01)
## annotation all the variables
## gene regions:
tx.file = system.file('extdata', 'hg19_gencode_genesymbol.bed.txt', package = 'methylKit')
tx = fread(tx.file)
tx2 = tx[, 1:4]
colnames(tx2) = paste0('gene.', c('chr', 'start', 'end', 'tag'))
tx2
rm(tx)
## myDiff25p: annotate by single CpG DMR of gene parts
diffAnn = annotateWithGeneParts(as(myDiff0p, "GRanges"), gene.obj)
diffanno.tss = as.data.table([email protected])
xx = as.data.table(as.data.frame([email protected], col.names=colnames(myDiff0p)))
diffanno.tss = cbind(diffanno.tss, xx[diffanno.tss$target.row,])
diffanno.tss = merge(diffanno.tss, tx2, by.x = 'feature.name', by.y = 'gene.tag', all.x = T, all.y = F)
diffanno.tss = diffanno.tss[!is.na(meth.diff),]
diffanno.tss
getTargetAnnotationStats(diffAnn, percentage=TRUE, precedence=TRUE)
## annotate by DMR of CpG locations
diffCpGann = annotateWithFeatureFlank(as(myDiff0p,"GRanges"), cpg.obj$CpGi, cpg.obj$shores, feature.name = "CpGi", flank.name = "shores")
getFeatsWithTargetsStats(diffCpGann,percentage=TRUE)
diffCpGanno.v2 = cbind(xx, diffCpGann@members)
diffCpGanno.v2
## annotate myDiff25p.promoter
xx = as.data.table(as.data.frame([email protected], col.names=colnames(myDiff.promoter.region)))
xx[, tag := paste0(chr, ':', start)]
tmp = tx2[, tag := paste0(gene.chr, ':', gene.start)]
myDiff.promoter = merge(xx, tmp, by.x='tag', by.y='tag')
dim(myDiff.promoter[pvalue < 0.01 & qvalue < 0.01,])
## annotation of cpgi region
cpg.file = system.file('extdata', 'hg19_cpg_island_tag_genesymbol.bed.txt', package = 'methylKit')
cpg.file
cpg = fread(cpg.file)
head(cpg)
colnames(cpg) = paste0('cpg.', c('chr', 'start', 'end', 'tag'))
cpg[, tag := paste0(cpg.chr, ':', cpg.start)]
xx = as.data.table(as.data.frame([email protected], col.names=colnames(myDiff.cpgi.region)))
xx[, type := 'cpgi']
xx2 = as.data.table(as.data.frame([email protected], col.names=colnames(myDiff.cpgi.shore)))
xx2[, type := 'shore']
rbind(xx, xx2) -> xx3
cpgi.region.shore = xx3
rm(xx, xx2, xx3)
xx = annotateWithGeneParts(as(cpgi.region.shore, "GRanges"), gene.obj)
as.data.table([email protected]) -> xx.tss
cpgi.region.shore.anno = cbind(cpgi.region.shore[xx.tss$target.row,], xx.tss)
cpgi.region.shore.anno[pvalue < 0.01 & qvalue < 0.01,]
table(sign(cpgi.region.shore.anno[pvalue < 0.01 & qvalue < 0.01, meth.diff]))
diffanno.tss # single CpG associated with TSS
diffCpGanno.v2 # single CpG associated with CpG island or shore
myDiff.promoter # promoter region count, DMR, and associated with gene symbol
cpgi.region.shore.anno # cpgi region count, DMR, and
options(scipen = 10)
diffanno.tss[grep('GATA3', feature.name),]
fwrite(diffanno.tss, file=paste0(cfg$methylDir, '/diff_cpg_genepart_anno_tss.tsv'), sep="\t", quote=F)
fwrite(diffCpGanno.v2, file=paste0(cfg$methylDir, '/diff_cpg_cpg_shore_anno.tsv'), sep="\t", quote=F)
fwrite(myDiff.promoter, file=paste0(cfg$methylDir, '/diff_promoter_region_anno.tsv'), sep="\t", quote=F)
fwrite(cpgi.region.shore.anno, file=paste0(cfg$methylDir, '/diff_cpg_island_region_shore_anno.tsv'), sep="\t", quote=F)
# diffanno.tss.gr = as(diffanno.tss, "GRanges")
# diffCpGanno.gr = as(diffCpGanno.v2, "GRanges")
# myDiff.promoter.gr = as(myDiff.promoter, "GRanges")
# cpgi.region.shore.anno.gr = as(cpgi.region.shore.anno, "GRanges")
tx = fread(tx.file)
tx3 = tx[, 1:8]; tx3
tx3[, symbol := sub(".*_", "", V4)]
colnames(tx3) = c('chr', 'start', 'end', 'tag', 'score', 'strand', 'cds.start', 'cds.end', 'symbol')
tx3[symbol != 'NA' & symbol != '',]
tx.gr = as(tx3, 'GRanges')
tx.gr
rm(tx3, tx)
save(tx.gr, diffanno.tss, diffCpGanno.v2, myDiff.promoter, cpgi.region.shore.anno, file='last4.Rdata') ##
# prepare for plotting specific region
con = dbConnect(drv=SQLite(), dbname=paste0(cfg$methylDir, '/sqlite/methyl.raw.sqlite'))
head(cpg)
plot.cpg = function(gene){
g.cpgi = cpg[grep(gene, cpg.tag),]
if(nrow(g.cpgi) < 1) { return(paste0('no cpgi found for gene ', gene))
for(i in
g.bmin = g.cpgi$cpg.start - 1
g.bmax = g.cpgi$cpg.end + 1
g.chr = g.cpgi$cpg.chr
g.bmin;g.bmax;g.chr
g.cpg.sel = dbGetQuery(con, paste0('select * from methylRaw where coverage > 10 and chr="', g.chr, '" and base > ', g.bmin, ' and base < ', g.bmax))
g.cpg.sel = as.data.table(g.cpg.sel)
g.cpg.sel[, t12:= 'T1']
g.cpg.sel[grep('T2', samplename), t12:= 'T2']
g.cpg.sel[, tag := paste0(chr, ':', base)]
g.cpg.sel
sig.cpg = diffanno.tss[chr == g.chr & start > g.bmin & end < g.bmax & qvalue < 0.01 & abs(meth.diff) > 25, tag]
sig.cpg
g.cpg.sel[tag %in% sig.cpg,]
gg = ggplot(g.cpg.sel, aes(x = base, y = samplename, color = freqC)) +
geom_point(size=3) + xlab('') + scale_colour_gradient(high = "#132B43", low = "#56B1F7")
ggsave(gg, file=paste0(cfg$methylDir, '/gene_gata3_CpG509.pdf'), width=29, height=2)
}
sync('r_fang/methyl')
## ma plot
tmp = diffanno.tss[abs(dist.to.feature) < 2000, ]
tmp[, Significance := 'N.S.']
tmp[qvalue < 0.001 & abs(meth.diff) > 50, Significance := 'Sig.']
table(tmp$Significance)
tmp
gg = ggplot(tmp, aes(x = meth.diff, y = -log(pvalue) + 3, color = Significance)) +
geom_point(size=1) + xlab("Methylation level (T2 vs T1)") + ylab("-log10(pvalue)")
ggsave(gg, file=paste0(cfg$methylDir, '/maplot.png'), width=4, height=3)
sync('r_fang/methyl')
dbDisconnet(con)
ls()
xxa = xx[1:10,]
xxa
xxa[, sampleid := sub(".*(\\d+)$", "\\1", variable)]
value.var = colnames(meth.dt)[5:ncol(meth.dt)]
merge(meth.dt, value.var = colnames(meth.dt)[5:ncol(meth.dt)])
dcast(meth.dt)
## convert obj to db
cfg$methylDir
objDB = makeMethylDB(methyl.fil.o, 'r_fang/methyl/exMethDB')
library(GenomicRanges)
my.win=GRanges(seqnames="chr21", ranges=IRanges(start=seq(from=9764513,by=10000,length.out=20),width=5000) )
# selects the records that lie inside the regions
selectByOverlap(myobj[[1]], my.win)
perc.meth = percMethylation(meth)
library(methylKit)
pdf(file=paste0(cfg$methylDir, '/mix_model_cost.pdf'))
myMixdml = myDiff.to.mixmdl(myDiff, plot=T, main='SCC T2 vsT1')
plotCost(myMixmdl, main='cost function')
dev.off()
save.image()
## END
### https://bioconductor.org/packages/release/bioc/vignettes/methylKit/inst/doc/methylKit.html
## generate the alignment file
targets = fread(mapfile, header = F)
## group == library
targets = setNames(targets, c('group', 'sampleName', 'runID', 'fastqDir', 'seqType')) #
targets.aln = copy(targets)
targets.aln[, fastq1 := '']
targets.aln[, fastq2 := '']
targets.aln[, fastq1.full := '']
targets.aln[, fastq2.full := '']
targets.aln = targets.aln[0,]
targets.aln
for(i in 1:nrow(targets)){
tmp = copy(targets[i,])
fastqs = dir(path=tmp$fastqDir, pattern="*.fastq.gz")
fastq1 = fastqs[grep("_R1_", fastqs)]
tmp = tmp[rep(1,length(fastq1)),]
tmp[, fastq1 := fastq1]
tmp[, fastq2 := sub("_R1_", "_R2_", fastq1)]
tmp[, fastq1.full := paste0(fastqDir, '/', fastq1)]
tmp[, fastq2.full := paste0(fastqDir, '/', fastq2)]
if(! (all(file.exists(tmp$fastq1.full)) & all(file.exists(tmp$fastq2.full)))){
cat("fastq file name problem:\n", file = logfile, append=T)
write.table(tmp, logfile, append=T)
break()
}
rbind(targets.aln, tmp) -> targets.aln
}
targets.aln
## align BWA
my $r1adaptor = 'AGATCGGAAGAGCACACGTCT';
my $r2adaptor = 'AGATCGGAAGAGCGTCGTGTA';
cwd = getwd()
## sample group runID
targets.aln[, sgrTag := paste0(sampleName, '_', group, '_', runID)]
targets.aln[, sgrDir := paste0(resDir, '/', sgrTag)]
targets.aln[, {system(paste0('mkdir -p ', sgrDir))}, by = 1:nrow(targets.aln)]
## zcat fastq
targets.aln[, fastq1.zcatfile := file.path(sgrDir, sub(".gz", "", fastq1))]
targets.aln[, fastq2.zcatfile := file.path(sgrDir, sub(".gz", "", fastq2))]
targets.aln[, zcat1.jobname := paste0("zcat.", sampleName, '.', fastq1)]
targets.aln[, zcat1.cmd := bsub.head(zcat1.jobname, mem=1, cpu=1, We='6:59', cwd)]
targets.aln[, zcat1.cmd:= paste0(zcat1.cmd, ' "/bin/zcat ', fastq1.full, ' > ', fastq1.zcatfile, '"')]
targets.aln[, zcat2.jobname := paste0("zcat.", sampleName, '.', fastq2)]
targets.aln[, zcat2.cmd := bsub.head(zcat2.jobname, mem=1, cpu=1, We='6:59', cwd)]
targets.aln[, zcat2.cmd := paste0(zcat2.cmd, ' "/bin/zcat ', fastq2.full, ' > ', fastq2.zcatfile, '"')]
targets.aln$zcat.cmd2[1]
exe.jobs(targets.aln$zcat1.cmd, cfg$logger)
exe.jobs(targets.aln$zcat2.cmd, cfg$logger)
## convert quality score
targets.aln[, fastq1.cqsfile := file.path(sgrDir, sub(".gz", ".cqs", fastq1))]
targets.aln[, fastq2.cqsfile := file.path(sgrDir, sub(".gz", ".cqs", fastq2))]
targets.aln$fastq1.cqsfile[1]
targets.aln[, cqs1.jobname := paste0("cqs1.", sampleName, '.', fastq1)]
targets.aln[, cqs1.cmd := bsub.head(cqs1.jobname, mem=1, cpu=1, We='2:59', cwd, zcat1.jobname), by=1:nrow(targets.aln)]
targets.aln[, cqs1.cmd := paste0(cqs1.cmd, ' "', ConvertQualityScore, ' --input ', fastq1.zcatfile, ' --output ', fastq1.cqsfile, '"')]
targets.aln$cqs1.cmd[1]
targets.aln[, cqs2.jobname := paste0("cqs2.", sampleName, '.', fastq1)]
targets.aln[, cqs2.cmd := bsub.head(cqs2.jobname, mem=1, cpu=1, We='2:59', cwd, zcat2.jobname), by = 1:nrow(targets.aln)]
targets.aln[, cqs2.cmd := paste0(cqs2.cmd, ' "', ConvertQualityScore, ' --input ', fastq2.zcatfile, ' --output ', fastq2.cqsfile, '"')]
targets.aln$cqs2.cmd[1]
exe.jobs(targets.aln$cqs2.cmd, cfg$logger)
exe.jobs(targets.aln$cqs2.cmd, cfg$logger)
## discard read less than half length
## find the read length
read.len = function(x){
gzfile(x, open='r') -> fh
readLines(fh, 2) -> h
nchar(h[2]) }
targets.aln[, read.len := sapply(targets.aln$fastq1.full, read.len)]
targets.aln$read.len
targets.aln[, fastq1.cutadaptfile := file.path(sgrDir, sub(".gz", ".cutadapt", fastq1))]
targets.aln[, fastq2.cutadaptfile := file.path(sgrDir, sub(".gz", ".cutadapt", fastq2))]
targets.aln$fastq1.cutadaptfile[1]
targets.aln[, cutadapt.jobname := paste0("cutadapt.", sgrTag)]
targets.aln[, cutadapt.cmd := bsub.head(cutadapt.jobname, mem=1, cpu=1, We='2:59', cwd, postdone = paste(cqs1.jobname, cqs2.jobname, collapse=" ")), by = 1:nrow(targets.aln)]
targets.aln[, cutadapt.cmd := paste(cutadapt.cmd, ' "', PYTHON, CUTADAPT, '-f fastq -a', clipR1, '-A', clipR2, '--overlap 10 --minimum-length', floor(read.len/2))]
targets.aln[, cutadapt.cmd := paste(cutadapt.cmd, '--quality-cutoff', bqtrim, '-o', fastq1.cutadaptfile, '--paired-output', fastq2.cutadaptfile, fastq1.cqsfile, fastq2.cqsfile, ' "')]
targets.aln$cutadapt.cmd[1]
exe.jobs(targets.aln$cutadapt.cmd, logger)
## breakpoint
wait4jobs(targets.aln$cutadapt.jobname, logger)
testOutputFiles = c(targets.aln$fastq1.cutadapt, targets.aln$fastq2.cutadapt)
if(delFiles){
del.files(targets.aln$fastq1.zcatfile, logger)
del.files(targets.aln$fastq2.zcatfile, logger)
del.files(targets.aln$fastq1.cqsfile, logger)
del.files(targets.aln$fastq2.cqsfile, logger)
}
## alignment
## targets.aln[, sgrTag := paste0(sampleName, '_', group, '_', runID)]
targets.aln[, reference := B37_BWA_INDEX]
targets.aln[, readgroup := paste0("@RG\\\\tID:", sgrTag, '_', seqType, '\\\\tPL:Illumina\\\\tPU:', sgrTag, "\\\\tLB:", sampleName, '_', group, "\\\\tSM:", sampleName)]
targets.aln[, bamfile := paste0(sgrDir, '/', sgrTag, '_', fastq1, '.bam')] #
targets.aln$bamfile[1]
file.exists(targets.aln$bamfile)
file.exists(targets.aln$fastq2.cutadaptfile)
targets.aln[, bwa.jobname := paste0("bwa.", sgrTag, fastq1)]
targets.aln[, bwa.cmd := bsub.head(bwa.jobname, mem=15, cpu=12, We='2:59', cwd)]
targets.aln[, bwa.cmd := paste0(bwa.cmd, ' "', bwa, ' mem -t 8 -T 0 -R \\\"', readgroup, '\\\" ', reference, ' ', fastq1.cutadaptfile, ' ', fastq2.cutadaptfile, ' | ', cfg$samtools, ' view -Shb -o ', bamfile, ' - "')]
targets.aln$bwa.cmd[1]
exe.jobs(targets.aln$bwa.cmd, logger)
## breakpoint
wait4jobs(targets.aln$bwa.jobname, logger)
## clean files
testInputFiles = c(targets.aln$fastq1.cutadapt, targets.aln$fastq2.cutadapt)
testOutputFiles = targets.aln$bamFile
if(delFiles & all(file.size(testOutputFiles) > 2000000)){
del.files(targets.aln$fastq1.cutadapt, logger)
del.files(targets.aln$fastq2.cutadapt, logger)
}
## merge bam file for the same library/group
## may several r1/r2 fastq files
tmpdir='/scratch/huw/'
targets.bam = targets.aln[, .(bamFileList = paste(paste0(' I=', bamfile), collapse = ' '), bwaJobList = paste(bwa.jobname, collapse = ' ')), by = list(sampleName, group)]
targets.bam[, libMergedBamFile := paste0(resDir, '/', sampleName, '_library', group, '_merged.bam')]
targets.bam[, merge.jobname := paste0('mergebam.', sampleName, group)]
targets.bam[, merge.cmd := bsub.head(merge.jobname, mem=30, cpu=8, We='2:26', cwd=cwd, postdone = bwaJobList)]
targets.bam[, merge.cmd := paste0(merge.cmd, ' "', JAVA, '/java -Xms256m -Xmx30g -Djava.io.tmpdir=', cfg$tmpdir, ' -jar ', cfg$PICARD, '/picard.jar MergeSamFiles ', bamFileList, ' O=', libMergedBamFile, ' SORT_ORDER=coordinate VALIDATION_STRINGENCY=LENIENT TMP_DIR=', tmpdir, ' CREATE_INDEX=TRUE USE_THREADING=FALSE MAX_RECORDS_IN_RAM=5000000 "')]
targets.bam$merge.cmd[2]
exe.jobs(targets.bam$merge.cmd, logger)
## markDuplicates for reads from the same library/group
## merge bam file from the same library/group
## output in resDir (r_fang)
targets.bam[, mkdupedLibBamFile := paste0(resDir, '/', sampleName, '_library', group, '_merged_mkdup.bam')]
targets.bam[, libmkdupMetricsFile := paste0(matricsDir, '/', sampleName, '_library', group, '_merged_mkdup_metrics.txt')]
targets.bam[, mkdup.jobname := paste0('mkdup.', sampleName, '.', group)]
targets.bam[, mkdup.cmd := bsub.head(mkdup.jobname, mem=30, cpu=3, We='8:26', cwd=cwd, postdone = merge.jobname)]
targets.bam[, mkdup.cmd := paste0(mkdup.cmd, ' "', JAVA, '/java -Xms256m -Xmx30g -Djava.io.tmpdir=', tmpdir, ' -jar ', cfg$PICARD, '/picard.jar MarkDuplicates I=', libMergedBamFile, ' O=', mkdupedLibBamFile, ' METRICS_FILE=', libmkdupMetricsFile, ' VALIDATION_STRINGENCY=LENIENT TMP_DIR=', tmpdir, ' REMOVE_DUPLICATES=TRUE CREATE_INDEX=TRUE MAX_RECORDS_IN_RAM=5000000 "')]
targets.bam$mkdup.cmd[3]
file.exists(targets.bam$libMergedBamFile)
exe.jobs(targets.bam$mkdup.cmd, logger)
## merge bam file for each sample
tmpdir='/scratch/huw/'
targets.smpBam = targets.bam[, .(BamFileList = paste(paste0(' I=', mkdupedLibBamFile), collapse = ' '), mkdupJobList = paste(mkdup.jobname, collapse = ' '), bamFile = paste(mkdupedLibBamFile, collapse=' '), fileNumber = length(mkdupedLibBamFile)), by = list(sampleName)]
colnames(targets.smpBam)
targets.smpBam
targets.smpBam[, smpBamFile := paste0(resDir, '/', sampleName, '_merged_mkdup_smp.bam')]
targets.smpBam[, smp.jobname := paste0('smp.merge.', sampleName)]
targets.smpBam$smp.jobname
targets.smpBam[, smp.cmd := bsub.head(smp.jobname, mem=30, cpu=8, We='8:26', cwd=cwd, postdone = mkdupJobList)]
targets.smpBam[, smp.cmd := paste0(smp.cmd, ' "', JAVA, '/java -Xms256m -Xmx30g -Djava.io.tmpdir=', tmpdir, ' -jar ', cfg$PICARD, '/picard.jar MergeSamFiles ', BamFileList, ' TMP_DIR=', tmpdir, ' DIR=', tmpdir, ' O=', smpBamFile, ' SORT_ORDER=coordinate VALIDATION_STRINGENCY=LENIENT TMP_DIR=', tmpdir, ' CREATE_INDEX=TRUE USE_THREADING=FALSE MAX_RECORDS_IN_RAM=5000000 "')]
targets.smpBam$smp.cmd[3]
exe.jobs(targets.smpBam$smp.cmd, logger)
## breakpoint before collect metrics
wait4jobs(targets.smpBam$smp.jobname, logger)
## clean files
testInputFiles = targets.smpBam$bamFile
testOutputFiles = targets.smpBam$smpBamFile
if(delFiles & all(file.size(testOutputFiles) > 2000000)){
del.files(targets.aln$fastq1.cutadapt, logger)
}
#else{stop("something might wrong here: the _merged_mkdup_smp.bam file for sample dont have the right size\n")}
## calculate HS matrix for the bam file for each sample
## the baits targets and the bam file
## variant_pipeline: 1402
tmpdir='/scratch/huw/'
targets.smpBam[, hmsFile := paste0(matricsDir, '/', 'hms_', sampleName, '_HsMatrix.txt')]
targets.smpBam[, hms.jobname := paste0('hms.', sampleName)]
targets.smpBam[, hms.cmd := bsub.head(hms.jobname, mem=10, cpu=1, We='8:26', cwd=cwd, postdone = smp.jobname)]
targets.smpBam[, hms.cmd := paste0(hms.cmd, ' "', JAVA, '/java -Xms256m -Xmx30g -Djava.io.tmpdir=', tmpdir, ' -jar ', PICARD, '/picard.jar CalculateHsMetrics I=', smpBamFile, ' TMP_DIR=', tmpdir, ' O=', hmsFile, ' REFERENCE_SEQUENCE=', genomeFasta, ' METRIC_ACCUMULATION_LEVEL=SAMPLE BAIT_INTERVALS=', baits_ilist, ' BAIT_SET_NAME=', wesImpactTarget, ' TARGET_INTERVALS=', targets_ilist, ' VALIDATION_STRINGENCY=LENIENT "')]
targets.smpBam$hms.cmd[3]
exe.jobs(targets.smpBam$hms.cmd, logger)
## for pair end, calculate the insert size
tmpdir='/scratch/huw/'
targets.smpBam[, insertSizeFile := paste0(matricsDir, '/', 'readlen_', sampleName, '_insertSizeFile.txt')]
targets.smpBam[, insertSizeHistgramFile := paste0(matricsDir, '/', 'readlen_', sampleName, '_insertSizeFile.txt')]
targets.smpBam[, readlen.jobname := paste0('readlen.', sampleName)]
targets.smpBam[, readlen.cmd := bsub.head(readlen.jobname, mem=10, cpu=1, We='8:26', cwd=cwd, postdone = smp.jobname)]
targets.smpBam[, readlen.cmd := bsub.head(readlen.jobname, mem=10, cpu=1, We='8:26', cwd=cwd)]
targets.smpBam[, readlen.cmd := paste0(readlen.cmd, ' "', JAVA, '/java -Xms256m -Xmx10g -Djava.io.tmpdir=', tmpdir, ' -jar ', PICARD, '/picard.jar CollectInsertSizeMetrics I=', smpBamFile)]
targets.smpBam[, readlen.cmd := paste0(readlen.cmd, ' O=', insertSizeFile, ' REFERENCE_SEQUENCE=', genomeFasta, ' METRIC_ACCUMULATION_LEVEL=SAMPLE HISTOGRAM_FILE=', insertSizeHistgramFile)]
targets.smpBam[, readlen.cmd := paste0(readlen.cmd, ' VALIDATION_STRINGENCY=LENIENT TMP_DIR=', tmpdir, ' "')]
targets.smpBam$readlen.cmd[3]
exe.jobs(targets.smpBam$readlen.cmd, logger)
## CollectAlignmentSummaryMetrics
tmpdir='/scratch/huw/'
targets.smpBam[, alnSummaryFile := paste0(matricsDir, '/', 'alnSum_', sampleName, '_alnSummaryFile.txt')]
targets.smpBam[, alnSum.jobname := paste0('alnSum.', sampleName)]
targets.smpBam[, alnSum.cmd := bsub.head(alnSum.jobname, mem=10, cpu=1, We='8:26', cwd=cwd, postdone = smp.jobname)]
#targets.smpBam[, alnSum.cmd := bsub.head(alnSum.jobname, mem=10, cpu=1, We='8:26', cwd=cwd)]
targets.smpBam[, alnSum.cmd := paste0(alnSum.cmd, ' "', JAVA, '/java -Xms256m -Xmx30g -Djava.io.tmpdir=', tmpdir, ' -jar ', PICARD, '/picard.jar CollectAlignmentSummaryMetrics I=', smpBamFile)]
targets.smpBam[, alnSum.cmd := paste0(alnSum.cmd, ' O=', alnSummaryFile, ' REFERENCE_SEQUENCE=', genomeFasta, ' METRIC_ACCUMULATION_LEVEL=SAMPLE TMP_DIR=', tmpdir, ' "')]
targets.smpBam$alnSum.cmd[3]
exe.jobs(targets.smpBam$alnSum, logger)
## CollectOxoGMetrics
tmpdir='/scratch/huw/'
targets.smpBam[, OxoFile := paste0(matricsDir, '/', 'Oxo_', sampleName, '_OxoFile.txt')]
targets.smpBam[, Oxo.jobname := paste0('Oxo.', sampleName)]
targets.smpBam[, Oxo.cmd := bsub.head(Oxo.jobname, mem=4, cpu=1, We='8:26', cwd=cwd, postdone = smp.jobname)]
targets.smpBam[, Oxo.cmd := bsub.head(Oxo.jobname, mem=4, cpu=1, We='8:26', cwd=cwd)]
targets.smpBam[, Oxo.cmd := paste0(Oxo.cmd, ' "', JAVA, '/java -Xms256m -Xmx4g -Djava.io.tmpdir=', tmpdir, ' -jar ', PICARD, '/picard.jar CollectOxoGMetrics I=', smpBamFile)]
targets.smpBam[, Oxo.cmd := paste0(Oxo.cmd, ' REFERENCE_SEQUENCE=', genomeFasta, ' O=', OxoFile, ' DB_SNP=', DB_SNP, ' VALIDATION_STRINGENCY=LENIENT')];
targets.smpBam[, Oxo.cmd := paste0(Oxo.cmd,' TMP_DIR=', tmpdir, ' "')];
targets.smpBam$Oxo.cmd[3]
exe.jobs(targets.smpBam$Oxo.cmd, logger)
## depth of coverage
tmpdir='/scratch/huw/'
targets.smpBam[, depthCoverageFile := paste0(matricsDir, '/', 'depthCoverage_', sampleName, '_depthCoverageFile.txt')]
targets.smpBam[, depthCoverage.jobname := paste0('depthCoverage.', sampleName)]
targets.smpBam[, depthCoverage.cmd := bsub.head(depthCoverage.jobname, mem=4, cpu=1, We='8:26', cwd=cwd, postdone = smp.jobname)]
targets.smpBam[, depthCoverage.cmd := bsub.head(depthCoverage.jobname, mem=4, cpu=1, We='8:26', cwd=cwd)]
targets.smpBam[, depthCoverage.cmd := paste0(depthCoverage.cmd, ' "', JAVA, '/java -Xms256m -Xmx30g -Djava.io.tmpdir=', tmpdir, ' -jar ', GATK, '/GenomeAnalysisTK.jar -T DepthOfCoverage')]
targets.smpBam[, depthCoverage.cmd := paste0(depthCoverage.cmd, ' -I ', smpBamFile, ' -R ', genomeFasta, ' -o ', depthCoverageFile, ' -L ', FP_INT, ' -rf BadCigar')]
targets.smpBam[, depthCoverage.cmd := paste0(depthCoverage.cmd, ' -mmq 20 -mbq 0 -omitLocusTable -omitSampleSummary -baseCounts --includeRefNSites -omitIntervals "')]
targets.smpBam$depthCoverage.cmd[3]
exe.jobs(targets.smpBam$depthCoverage.cmd, logger)
## GC bias metrics
tmpdir = "/scratch/huw/"
targets.smpBam[, gcbiasFile := paste0(matricsDir, '/', 'gcbias_', sampleName, '_gcbiasFile.txt')]
targets.smpBam[, gcbiasSummaryFile := paste0(matricsDir, '/', 'gcbias_', sampleName, '_gcbiasSummaryFile.txt')]
targets.smpBam[, gcbiasChartFile := paste0(matricsDir, '/', 'gcbias_', sampleName, '_gcbiasChartFile.pdf')]
targets.smpBam[, gcbias.jobname := paste0('gcbias.', sampleName)]
targets.smpBam[, gcbias.cmd := bsub.head(gcbias.jobname, mem=4, cpu=1, We='8:26', cwd=cwd, postdone = smp.jobname)]
targets.smpBam[, gcbias.cmd := bsub.head(gcbias.jobname, mem=4, cpu=1, We='8:26', cwd=cwd)]
targets.smpBam[, gcbias.cmd := paste0(gcbias.cmd, ' "', JAVA, '/java -Xms256m -Xmx4g -Djava.io.tmpdir=', tmpdir, ' -jar ', PICARD, '/picard.jar CollectGcBiasMetrics I=', smpBamFile)]
targets.smpBam[, gcbias.cmd := paste0(gcbias.cmd, ' O=', gcbiasFile, ' REFERENCE_SEQUENCE=', genomeFasta, ' SUMMARY_OUTPUT=', gcbiasSummaryFile)]
targets.smpBam[, gcbias.cmd := paste0(gcbias.cmd, ' CHART_OUTPUT=', gcbiasChartFile, ' TMP_DIR=', tmpdir, ' "')]
targets.smpBam$gcbias.cmd[3]
exe.jobs(targets.smpBam$gcbias.cmd, logger)
## summarize metrics file
## need add
## breakpoint after collect metrics
testJobname = c(targets.smpBam$gcbias.jobname, targets.smpBam$Oxo.jobname, targets.smpBam$depthCoverage.jobname, targets.smpBam$alnSum.jobname, targets.smpBam$readlen.jobname, targets.smpBam$hms.jobname)
wait4jobs(testJobname, logger)
## cleaning and realignment
## organize the targets
targets.smpGroup = fread(groupfile, header = F)
targets.smpGroup = setNames(targets.smpGroup, c('sampleName', 'smpGroup'))
targets.smpGroup
targets.smpGroup = merge(targets.smpBam[, .(sampleName, smpBamFile, smp.jobname)], targets.smpGroup, by = 'sampleName')
targets.smpGroup
targets.recal = targets.smpGroup[, .(bamFileList = paste(paste0('-I ', smpBamFile), collapse=' '), smpJobnameList = paste(smp.jobname, collapse = ', ')), by=smpGroup]
targets.recal
## REALIGNTARGETCREATOR
tmpdir = "/scratch/huw/"
targets.recal[, realnGenTargetFile := paste0(matricsDir, '/', 'realn_',smpGroup, '_realnGenTargetFile.intervals')]
targets.recal[, realnGen.jobname := paste0('realnGen_', smpGroup)]
targets.recal[, realnGen.cmd := bsub.head(realnGen.jobname, mem=15, cpu=1, We='8:26', cwd=cwd, postdone = smpJobnameList)]
targets.recal[, realnGen.cmd := bsub.head(realnGen.jobname, mem=15, cpu=1, We='8:26', cwd=cwd)]
targets.recal[, realnGen.cmd := paste0(realnGen.cmd, ' "', JAVA, '/java -Xms256m -Xmx15g -XX:-UseGCOverheadLimit -Djava.io.tmpdir=', tmpdir, ' -jar ', GATK)]
targets.recal[, realnGen.cmd := paste0(realnGen.cmd, '/GenomeAnalysisTK.jar -T RealignerTargetCreator -R ', genomeFasta, ' -known ', DB_SNP, ' -known ', MILLS_1000G, ' ', bamFileList)]
## keep the same the -known with the next step
targets.recal[, realnGen.cmd := paste0(realnGen.cmd, ' --out ', realnGenTargetFile, ' "')]
targets.recal$realnGen.cmd[1]
exe.jobs(targets.recal$realnGen.cmd, logger)
for(i in 1:nrow(targets.recal)){system(targets.recal$realnGen.cmd[i])}
targets.smpGroup$realnGenTargetFile = targets.recal[targets.smpGroup$smpGroup, realnGenTargetFile]
targets.smpGroup
## IndelRealigner
tmpdir = "/scratch/huw/"
targets.recal[, indelRealnOutFile := paste0(matricsDir, '/', 'realn_',smpGroup, '_indelRealn_OutFile.txt')]
targets.recal[, realnIndel.jobname := paste0('realnIndel.', smpGroup)]
targets.recal[, realnIndel.cmd := bsub.head(realnIndel.jobname, mem=15, cpu=1, We='8:26', cwd=cwd, postdone = realnGen.jobname)]
targets.recal[, realnIndel.cmd := bsub.head(realnIndel.jobname, mem=15, cpu=1, We='8:26', cwd=cwd)]
targets.recal[, realnIndel.cmd := paste0(realnIndel.cmd, ' "', JAVA, '/java -Xms256m -Xmx15g -XX:-UseGCOverheadLimit -Djava.io.tmpdir=', tmpdir, ' -jar ', GATK, '/GenomeAnalysisTK.jar')]
targets.recal[, realnIndel.cmd := paste0(realnIndel.cmd, ' -T IndelRealigner -R ', genomeFasta, ' -known ', DB_SNP, ' -known ', MILLS_1000G, ' -targetIntervals ', realnGenTargetFile)]
targets.recal[, realnIndel.cmd := paste0(realnIndel.cmd, ' -L ', targets_bed)]
targets.recal[, realnIndel.cmd := paste0(realnIndel.cmd, ' --noOriginalAlignmentTags -S LENIENT --maxReadsForRealignment 500000 --maxReadsInMemory 3000000 --maxReadsForConsensuses 500000 -rf BadCigar ')]
targets.recal[, realnIndel.cmd := paste0(realnIndel.cmd, bamFileList, ' -nWayOut _realnIndel.bam "')] ## filename change here from .bam to _realnindel.bam
# bamFileList => ' -I file1.bam -I file2.bam'
# the output bamfiles will be ' -I file1_realnIndel.bam -I file2_realnIndel.bam'
targets.recal$realnIndel.cmd[1]
#for(i in 1:nrow(targets.recal)){system(targets.recal$realnIndel.cmd[i])}
exe.jobs(targets.recal$realnIndel.cmd, logger)
# generated by the above realnIndel step nWayOut
# they are under the cwd !!!
targets.smpGroup[, smpRealnBamFile := sub(".bam", "_realnIndel.bam", smpBamFile)]
targets.smpGroup[, smpRealnBaiFile := sub(".bam", "_realnIndel.bai", smpBamFile)]
targets.smpGroup$realnIndel.jobname = targets.recal[targets.smpGroup$smpGroup, realnIndel.jobname]
targets.smpGroup
wait4jobs(targets.recal$realnIndel.jobname, logger)
for(i in 1:length(targets.smpGroup$smpRealnBamFile)){ system(paste0('mv ', targets.smpGroup$smpRealnBamFile[i], ' ', resDir))}
for(i in 1:length(targets.smpGroup$smpRealnBamFile)){ system(paste0('mv ', targets.smpGroup$smpRealnBaiFile[i], ' ', resDir))}
file.exists(targets.smpGroup$smpRealnBamFile)
file.exists(targets.smpGroup$smpBamFile)
## Realign use -known
## while recal use --knownSites
## baserecal
targets.smpGroup[, baserecalOutFile := paste0(matricsDir, '/', 'realn_', sampleName, '_recal_OutFile.txt')]
targets.smpGroup[, baserecal.jobname := paste0('baserecal.', sampleName)]
#targets.smpGroup[, baserecal.cmd := bsub.head(baserecal.jobname, mem=40, cpu=12, We='8:26', cwd=cwd)] , postdone = realnIndel.jobname)]
targets.smpGroup[, baserecal.cmd := bsub.head(baserecal.jobname, mem=40, cpu=12, We='8:26', cwd=cwd)]
targets.smpGroup[, baserecal.cmd := paste0(baserecal.cmd, ' "', JAVA, '/java -Xms256m -Xmx30g -XX:-UseGCOverheadLimit -Djava.io.tmpdir=', tmpdir, ' -jar ', GATK, '/GenomeAnalysisTK.jar')]
targets.smpGroup[, baserecal.cmd := paste0(baserecal.cmd, ' -T BaseRecalibrator -R ', genomeFasta, ' --knownSites ', DB_SNP, ' --knownSites ', MILLS_1000G)]
targets.smpGroup[, baserecal.cmd := paste0(baserecal.cmd, ' --knownSites ', HAPMAP, ' --knownSites ', OMNI_1000G, ' --knownSites ', PHASE1_SNPS_1000G)]
targets.smpGroup[, baserecal.cmd := paste0(baserecal.cmd, ' --covariate ContextCovariate --covariate CycleCovariate --covariate QualityScoreCovariate --covariate ReadGroupCovariate -rf BadCigar --num_cpu_threads_per_data_thread 12 ')]
targets.smpGroup[, baserecal.cmd := paste0(baserecal.cmd, ' -S LENIENT -rf BadCigar --out ', baserecalOutFile, ' -I ', smpRealnBamFile, ' "')]
targets.smpGroup$baserecal.cmd[1]
targets.smpGroup$baserecal.cmd
exe.jobs(targets.smpGroup$baserecal.cmd, logger)
## printreads
targets.smpGroup[, printReadsBQSROutFile := paste0(matricsDir, '/', 'realn_', sampleName, '_bqsr_outfile.txt')]
targets.smpGroup[, printReadsBamOutFile := sub('.bam', "_recal.bam", smpRealnBamFile)]
targets.smpGroup[, printReads.jobname := paste0('printReads.', sampleName)]
targets.smpGroup[, printReads.cmd := bsub.head(printReads.jobname, mem=30, cpu=6, We='8:26', cwd=cwd, postdone = baserecal.jobname)]
targets.smpGroup[, printReads.cmd := bsub.head(printReads.jobname, mem=30, cpu=6, We='8:26', cwd=cwd)]
targets.smpGroup[, printReads.cmd := paste0(printReads.cmd, ' "', JAVA, '/java -Xms256m -Xmx30g -XX:-UseGCOverheadLimit -Djava.io.tmpdir=', tmpdir, ' -jar ', GATK, '/GenomeAnalysisTK.jar')]
targets.smpGroup[, printReads.cmd := paste0(printReads.cmd, ' -T PrintReads -R ', genomeFasta, ' --emit_original_quals -BQSR ', baserecalOutFile)]
targets.smpGroup[, printReads.cmd := paste0(printReads.cmd, ' --num_cpu_threads_per_data_thread 6 -rf BadCigar --downsampling_type NONE' )]
targets.smpGroup[, printReads.cmd := paste0(printReads.cmd, ' --out ', printReadsBamOutFile, ' -I ', smpRealnBamFile, ' "')]
targets.smpGroup$printReads.cmd[1]
exe.jobs(targets.smpGroup$printReads.cmd, logger)
## breakpoint before SNP calling
testJobname = c(targets.smpGroup$printReads.jobname, targets.smpGroup$baserecal.jobname, targets.smpGroup$realnindel.jobname)
wait4jobs(testJobname, logger)
testOutputFiles = c(targets.smpGroup$printReadsBamOutFile)
if(delFiles & file.size(testOutputFiles) > 20000000){
del.files(targets.smpGroup$smpBamFile, logger)
del.files(targets.smpRealnBamFile$smpBamFile, logger)
}
#else{stop("something might wrong here: _merged_mkdup_smp_realnIndel_recal.bam\n")}
## split targets.bed by chrom
bed = fread(targets5bp_bed)
chrs = unique(bed$V1)
for(i in chrs){
fwrite(bed[V1 == i,], quote=F, sep="\t", col.names=F, row.names=F, file=paste0(matricsDir, '/targets5bp_', i, '.bed' ))
}
## build pair
targets.pair = fread(pairfile, header=F)
targets.pair = setNames(targets.pair, c('normal', 'tumor'))
targets.pair[, normalBamFile := paste0(initDir, '/', normal, '/', normal, '_merged_mkdup_smp_recal_realnIndel.bam')]
targets.pair[, tumorBamFile := paste0(initDir, '/', tumor, '/', tumor, '_merged_mkdup_smp_recal_realnIndel.bam')]
file.exists(targets.pair$tumorBamFile)
file.exists(targets.pair$normalBamFile)
## build chrPair
targets.chrPair = targets.pair[rep(1:nrow(targets.pair), each=length(chrs)), ]
targets.chrPair[, chr := rep(chrs, times=nrow(targets.pair))]
targets.chrPair[, chr.bed := paste0(matricsDir, '/targets5bp_', chr, '.bed')]
targets.chrPair
## call haplocalltyper by chromosome
## https://software.broadinstitute.org/gatk/best-practices/workflow?=id=11145
## 4.0 version
## haplotypeCaller
## ImportGenomicsDB
## genotypeGVCFs
## VariantRecalibrator
## ApplyRecalibration
targets.germline = targets.chrPair[!duplicated(paste0(normal, chr.bed)),]
dim(targets.germline)
targets.germline[, haplocallGVCFChrFile := paste0(initDir, '/', normal, '/haplocall_', normal, '_', chr, '.g.vcf.gz')]
targets.germline[, haplocall.jobname := paste0('haplocall.', normal, '.', chr)]
targets.germline[, haplocall.cmd := bsub.head(haplocall.jobname, mem=30, cpu=10), by=1:nrow(targets.germline)]
targets.germline[, haplocall.cmd := paste0(haplocall.cmd, " \"", GATK4, " --java-options '-Xms256m -Xmx30g -XX:-UseGCOverheadLimit -Djava.io.tmpdir=", tmpdir, "' HaplotypeCaller -R ", genomeFasta, " -L ", chr.bed, " -I ", normalBamFile, " -O ", haplocallGVCFChrFile, " -ERC GVCF --dbsnp ", DB_SNP, " --max-reads-per-alignment-start 0 \"")]
targets.germline$haplocall.cmd[1]
exe.jobs(targets.germline$haplocall.cmd, logger)
file.exists(targets.germline$haplocallGVCFChrFile)
## combine haplocall g.vcf per sample
targets.germline[, {
chrgvcf = paste(paste(' --variant ', haplocallGVCFChrFile), collapse=' ')
haplocallGVCFFile = paste0(initDir, '/', normal, '/haplo_', .BY, '.g.vcf.gz')
haplocat.jobname = paste0('haplocat.', .BY)
pd = paste(haplocall.jobname, collapse=' ')
cmd = bsub.head(haplocat.jobname, mem=30, cpu=10, postdone = pd)