forked from dozmorovlab/PDXHiC_supplemental
-
Notifications
You must be signed in to change notification settings - Fork 0
/
00_preprocessing.Rmd
712 lines (645 loc) · 36 KB
/
00_preprocessing.Rmd
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
---
title: "Analysis"
author: "Mikhail Dozmorov"
date: "`r Sys.Date()`"
output:
pdf_document:
toc: no
html_document:
theme: cerulean
toc: yes
---
```{r setup, echo=FALSE, message=FALSE, warning=FALSE}
# Set up the environment
library(knitr)
opts_chunk$set(cache.path='cache/', fig.path='img/', cache=F, tidy=T, fig.keep='high', echo=F, dpi=100, warnings=F, message=F, comment=NA, warning=F, results='as.is', fig.width = 10, fig.height = 6) #out.width=700,
library(pander)
panderOptions('table.split.table', Inf)
set.seed(1)
library(dplyr)
options(stringsAsFactors = FALSE)
```
# Libraries
```{r libraries}
library(GenomicRanges)
library(GenomicInteractions)
library(InteractionSet)
library(data.table)
# library(MDmisc)
```
# Excludable regions
```{r}
if (!file.exists("excludable.rda")) {
library(rCGH)
library(AnnotationHub)
ah <- AnnotationHub()
# Centromeres
# Adjust chromosome names
hg38$chrom[hg38$chrom == 23] <- "X"
hg38$chrom[hg38$chrom == 24] <- "Y"
hg38$chrom <- paste0("chr", hg38$chrom)
# Make GRanges object
hg38.UCSC.centromere <- makeGRangesFromDataFrame(hg38, seqnames.field = "chrom", start.field = "centromerStart", end.field = "centromerEnd")
# Assign seqinfo data
seqlengths(hg38.UCSC.centromere) <- hg38$length
genome(hg38.UCSC.centromere) <- "hg38"
# Telomeres
query_data <- query(ah, c("excluderanges", "UCSC", "Homo Sapiens", "hg38", "telomere"))
hg38.UCSC.telomere <- query_data[["AH95938"]]
# Excluderanges
query_data <- query(ah, "excluderanges")
excludeGR.hg38.Kundaje.1 <- query_data[["AH95917"]]
save(hg38.UCSC.centromere, hg38.UCSC.telomere, excludeGR.hg38.Kundaje.1, file = "excludable.rda")
} else {
load("excludable.rda")
}
```
# Functions
- PR unique - green, 0,255,0
- CR unique - red, 255,0,0
- Common - blue, 0,0,255
```{r functions}
#' Convert GInteractions to BEDPE format, color according to condition
#' @param mtx_selected data frame from GInteractions
#' @param col color
#' @param conditon condition, goes into the name field
#' Returns a data frame with coordinates kept in full numerical format
toBEDPE <- function(mtx_selected = as.data.frame(loops_XX_common), col = "0,255,0", condition = "Common") {
x_selected <- data.frame(chr1 = as.character(mtx_selected$seqnames1),
x1 = mtx_selected$start1,
x2 = mtx_selected$end1,
chr2 = mtx_selected$seqnames2,
y1 = mtx_selected$start2,
y2 = mtx_selected$end2,
name = paste(condition, mtx_selected$name, sep = "_"),
score = ".",
strand1 = ".",
strand2 = ".",
color = col)
# Keep full numbers
x_selected$x1 <- format(x_selected$x1, scientific = FALSE, trim = TRUE, justify = "none")
x_selected$x2 <- format(x_selected$x2, scientific = FALSE, trim = TRUE, justify = "none")
x_selected$y1 <- format(x_selected$y1, scientific = FALSE, trim = TRUE, justify = "none")
x_selected$y2 <- format(x_selected$y2, scientific = FALSE, trim = TRUE, justify = "none")
return(x_selected)
}
#' Convert GRanges to BED format, color according to condition
#' @param mtx_selected data frame from GRanges
#' @param col color
#' @param conditon condition, goes into the name field
#' Returns a data frame with coordinates kept in full numerical format
toBED <- function(mtx_selected = as.data.frame(anchors_PR_common), col = "0,255,0", condition = "Common", header = FALSE) {
x_selected <- data.frame(chr = as.character(mtx_selected$seqnames),
start = mtx_selected$start,
end = mtx_selected$end,
name = condition,
score = mtx_selected$freq,
strand = ".",
thickStart = mtx_selected$start,
thickEnd = mtx_selected$end,
color = col)
# Largest on top
x_selected <- x_selected[order(x_selected$score, decreasing = TRUE), ]
# Keep full numbers
x_selected$start <- format(x_selected$start, scientific = FALSE, trim = TRUE, justify = "none")
x_selected$end <- format(x_selected$end, scientific = FALSE, trim = TRUE, justify = "none")
x_selected$thickStart <- format(x_selected$thickStart, scientific = FALSE, trim = TRUE, justify = "none")
x_selected$thickEnd <- format(x_selected$thickEnd, scientific = FALSE, trim = TRUE, justify = "none")
# Append header
if (header){
x_selected <- rbind(c("track itemRgb=On", rep("", ncol(x_selected) - 1)), x_selected)
}
return(x_selected)
}
#' Summary of loop width
#' @param loops GInteractions object
#' @return Formatted string of loop min/median/mean/max
width_summary <- function(prefix = "Loops PR all width", loops = loops_PR_unique) {
width_vector <- pairdist(loops)
# https://stackoverflow.com/questions/33047601/paste-collapse-with-tabs
width_string <- c(list(prefix), list(min(width_vector)), list(median(width_vector)), list(round(mean(width_vector), digits = 2)), list(max(width_vector)))
return(width_string)
}
#' Loop width difference test
#' @param loops1 first GInteraction object
#' @param loops2 second GInteraction object
#' @return Formatted Wilcoxon p-value of width distribution differences
width_difference <- function(loops1 = loops_PR_unique, loops2 = loops_CR_unique) {
width_vector1 <- pairdist(loops1)
width_vector2 <- pairdist(loops2)
# If width is NA (for neoloops), return 1
if (sum(is.na(width_vector1)) > 0 | sum(is.na(width_vector2)) > 0) {
return(1)
} else {
return(formatC(wilcox.test(width_vector1, width_vector2)$p.value, format = "e", digits = 2))
}
}
```
# Settings
```{r settings, echo=TRUE}
# Resolution, or Window size for SpectralTAD only
res <- "10"
# How to treat adjacent anchors
merge_adjacent <- TRUE
# Overlap type for findOverlaps
if (merge_adjacent) {
overlap_type <- "any"
} else {
overlap_type <- "equal"
}
```
# Load data
Creation of PR and CR GInteractions
## Mustache
```{r mustache, eval=FALSE}
# Mustache data
dir_data <- "/Users/mdozmorov/Documents/Data/GoogleDrive/HiC_files/results/mustache_results_v2"
fileNameIn1 <- file.path(dir_data, paste0("VC_UCD52PR_", res, ".tsv")) # PR
fileNameIn2 <- file.path(dir_data, paste0("VC_UCD52CR_", res, ".tsv")) # CR
# Function to construct GInteractions
# Should be modified for different data
mtx2GInteractions <- function(fileNameIn = fileNameIn1, condition = "PR") {
# Load data
mtx <- read.table(fileNameIn, sep = "\t", header = TRUE)
gr1 <- data.frame(chr = paste0("chr", mtx[, 1]),
start = mtx[, 2],
end = mtx[, 3]) %>% makeGRangesFromDataFrame(., seqinfo = unique(c(paste0("chr", mtx[, 1]), paste0("chr", mtx[, 4]))))
gr2 <- data.frame(chr = paste0("chr", mtx[, 4]),
start = mtx[, 5],
end = mtx[, 6]) %>% makeGRangesFromDataFrame(., seqinfo = unique(c(paste0("chr", mtx[, 1]), paste0("chr", mtx[, 4]))))
# Domain GRanges overlapping centromeres
# gr <- data.frame(chr = mtx[, 1],
# start = mtx[, 2],
# end = mtx[, 6]) %>% makeGRangesFromDataFrame(., seqinfo = unique(mtx[, 1]))
# gr_centromere <- findOverlaps(gr, hg38.UCSC.centromere, type = "any")
gr1_centromere <- findOverlaps(gr1, hg38.UCSC.centromere, type = "any")
gr2_centromere <- findOverlaps(gr2, hg38.UCSC.centromere, type = "any")
# Boundaries overlapping telomeres
gr1_telomeres <- findOverlaps(gr1, hg38.UCSC.telomere, type = "any")
gr2_telomeres <- findOverlaps(gr2, hg38.UCSC.telomere, type = "any")
# Boundaries overlapping excludable regions
gr1_exclude <- findOverlaps(gr1, excludeGR.hg38.Kundaje.1, type = "any")
gr2_exclude <- findOverlaps(gr2, excludeGR.hg38.Kundaje.1, type = "any")
# Index to exclude
index_exclude <- sort(unique(c(queryHits(gr1_centromere), queryHits(gr2_centromere), queryHits(gr1_telomeres), queryHits(gr2_telomeres), queryHits(gr1_exclude), queryHits(gr2_exclude))))
# Complement index to include
index_include <- setdiff(seq(1:length(gr1)), index_exclude)
# GInteractions excluding selected boundaries
gi <- GInteractions(gr1[index_include], gr2[index_include], mode = "strict")
# Construct name
gi$name <- paste(condition,
paste("Width", calculateDistances(gi), sep = ":"),
paste("FDR", formatC(mtx[index_include, 7], format = "e", digits = 2), sep = ":"),
paste("SCALE", round(mtx[index_include, 8], digits = 2), sep = ":"), sep = "_")
return(gi)
}
# Create condition-specific GInteractions
gi_PR <- mtx2GInteractions(fileNameIn = fileNameIn1, condition = "PR")
gi_CR <- mtx2GInteractions(fileNameIn = fileNameIn2, condition = "CR")
# Make seqlevels the same
seqlevels(gi_PR) <- seqlevels(gi_CR) <- unique(c(seqlevels(gi_PR), seqlevels(gi_CR)))
```
## HiCcompare
```{r HiCcompare, eval=FALSE}
# HiCcompare data
dir_data <- "/Users/mdozmorov/Documents/Data/GoogleDrive/HiC_files/results/HiCcompare_results_v2"
fileNameIn1 <- file.path(dir_data, paste0(res, "kb_glob_fdr_DN.bedpe")) # PR
fileNameIn2 <- file.path(dir_data, paste0(res, "kb_glob_fdr_UP.bedpe")) # CR
# Function to construct GInteractions
# Should be modified for different data
mtx2GInteractions <- function(fileNameIn = fileNameIn1, condition = "PR") {
# Load data
mtx <- read.table(fileNameIn, sep = "\t", header = FALSE)
gr1 <- data.frame(chr = mtx[, 1],
start = mtx[, 2],
end = mtx[, 3]) %>% makeGRangesFromDataFrame(., seqinfo = unique(c(mtx[, 1], mtx[, 4])))
gr2 <- data.frame(chr = mtx[, 4],
start = mtx[, 5],
end = mtx[, 6]) %>% makeGRangesFromDataFrame(., seqinfo = unique(c(mtx[, 1], mtx[, 4])))
# Domain GRanges overlapping centromeres
# gr <- data.frame(chr = mtx[, 1],
# start = mtx[, 2],
# end = mtx[, 6]) %>% makeGRangesFromDataFrame(., seqinfo = unique(mtx[, 1]))
# gr_centromere <- findOverlaps(gr, hg38.UCSC.centromere, type = "any")
gr1_centromere <- findOverlaps(gr1, hg38.UCSC.centromere, type = "any")
gr2_centromere <- findOverlaps(gr2, hg38.UCSC.centromere, type = "any")
# Boundaries overlapping telomeres
gr1_telomeres <- findOverlaps(gr1, hg38.UCSC.telomere, type = "any")
gr2_telomeres <- findOverlaps(gr2, hg38.UCSC.telomere, type = "any")
# Boundaries overlapping excludable regions
gr1_exclude <- findOverlaps(gr1, excludeGR.hg38.Kundaje.1, type = "any")
gr2_exclude <- findOverlaps(gr2, excludeGR.hg38.Kundaje.1, type = "any")
# Index to exclude
index_exclude <- sort(unique(c(queryHits(gr1_centromere), queryHits(gr2_centromere), queryHits(gr1_telomeres), queryHits(gr2_telomeres), queryHits(gr1_exclude), queryHits(gr2_exclude))))
# Complement index to include
index_include <- setdiff(seq(1:length(gr1)), index_exclude)
# GInteractions excluding selected boundaries
gi <- GInteractions(gr1[index_include], gr2[index_include], mode = "strict")
# Construct name
gi$name <- paste(condition,
paste("Width", calculateDistances(gi), sep = ":"),
paste("M", round(mtx[index_include, 7], digits = 2), sep = ":"),
paste("D", round(mtx[index_include, 8], digits = 2), sep = ":"),
paste("A", round(mtx[index_include, 9], digits = 2), sep = ":"),
paste("padj", formatC(mtx[index_include, 10], format = "e", digits = 2), sep = ":"), sep = "_")
return(gi)
}
# Create condition-specific GInteractions
gi_PR <- mtx2GInteractions(fileNameIn = fileNameIn1, condition = "PR")
gi_CR <- mtx2GInteractions(fileNameIn = fileNameIn2, condition = "CR")
# Make seqlevels the same
seqlevels(gi_PR) <- seqlevels(gi_CR) <- unique(c(seqlevels(gi_PR), seqlevels(gi_CR)))
```
## NeoLoopFinder
```{r neoloopfinder, eval=TRUE}
# Neoloopfinder data
# dir_data <- "/Users/mdozmorov/Documents/Data/GoogleDrive/HiC_files/results/neoloopfinder_results_v2"
dir_data <- "/Users/mdozmorov/Documents/Data/GoogleDrive/HiC_files/results/neoloopfinder_results_v2/10-06-2021-neoloopfinder-neoloops-Kavita"
fileNameIn1 <- file.path(dir_data, "PR.neoloopfinder.loops.txt") # PR
fileNameIn2 <- file.path(dir_data, "CR.neoloopfinder.loops.txt") # CR
# Function to construct GInteractions
# Should be modified for different data
mtx2GInteractions <- function(fileNameIn = fileNameIn1, condition = "PR") {
# Load data
mtx <- read.table(fileNameIn, sep = "\t", header = FALSE)
# Subset to neoloops, marked as "1" in the third position of assembly column
mtx <- mtx[sapply(mtx[, 7], function(x) strsplit(x, ",")[[1]][3]) == "1", ]
gr1 <- data.frame(chr = mtx[, 1],
start = mtx[, 2],
end = mtx[, 3]) %>% makeGRangesFromDataFrame(., seqinfo = unique(c(mtx[, 1], mtx[, 4])))
gr2 <- data.frame(chr = mtx[, 4],
start = mtx[, 5],
end = mtx[, 6]) %>% makeGRangesFromDataFrame(., seqinfo = unique(c(mtx[, 1], mtx[, 4])))
# Domain GRanges overlapping centromeres
# gr <- data.frame(chr = mtx[, 1],
# start = mtx[, 2],
# end = mtx[, 6]) %>% makeGRangesFromDataFrame(., seqinfo = unique(mtx[, 1]))
gr1_centromere <- findOverlaps(gr1, hg38.UCSC.centromere, type = "any")
gr2_centromere <- findOverlaps(gr2, hg38.UCSC.centromere, type = "any")
# Boundaries overlapping telomeres
gr1_telomeres <- findOverlaps(gr1, hg38.UCSC.telomere, type = "any")
gr2_telomeres <- findOverlaps(gr2, hg38.UCSC.telomere, type = "any")
# Boundaries overlapping excludable regions
gr1_exclude <- findOverlaps(gr1, excludeGR.hg38.Kundaje.1, type = "any")
gr2_exclude <- findOverlaps(gr2, excludeGR.hg38.Kundaje.1, type = "any")
# Index to exclude
index_exclude <- sort(unique(c(queryHits(gr1_centromere), queryHits(gr2_centromere), queryHits(gr1_telomeres), queryHits(gr2_telomeres), queryHits(gr1_exclude), queryHits(gr2_exclude))))
# Complement index to include
index_include <- setdiff(seq(1:length(gr1)), index_exclude)
# GInteractions excluding selected boundaries
gi <- GInteractions(gr1[index_include], gr2[index_include], mode = "strict")
# Construct name
gi$name <- paste(condition, paste("Width", calculateDistances(gi), sep = ":"), mtx[index_include, 7], sep = "_")
return(gi)
}
# Create condition-specific GInteractions
gi_PR <- mtx2GInteractions(fileNameIn = fileNameIn1, condition = "PR")
gi_CR <- mtx2GInteractions(fileNameIn = fileNameIn2, condition = "CR")
# Make seqlevels the same
seqlevels(gi_PR) <- seqlevels(gi_CR) <- unique(c(seqlevels(gi_PR), seqlevels(gi_CR)))
```
## GENOVA
```{r genova, eval=FALSE}
# Neoloopfinder data
dir_data <- "/Users/mdozmorov/Documents/Data/GoogleDrive/HiC_files/results/TAD_Boundaries/GENOVA_IS"
fileNameIn1 <- file.path(dir_data, paste0("UCD52PR_", res, "000.pairedbed")) # PR
fileNameIn2 <- file.path(dir_data, paste0("UCD52CR_", res, "000.pairedbed")) # CR
# Load PR data
mtx_PR <- read.table(fileNameIn1, sep = "\t", header = FALSE)
gr1 <- data.frame(chr = mtx_PR[, 1],
start = mtx_PR[, 2],
end = mtx_PR[, 3]) %>% makeGRangesFromDataFrame(., seqinfo = unique(c(mtx_PR[, 1], mtx_PR[, 4])))
gr2 <- data.frame(chr = mtx_PR[, 4],
start = mtx_PR[, 5],
end = mtx_PR[, 6]) %>% makeGRangesFromDataFrame(., seqinfo = unique(c(mtx_PR[, 1], mtx_PR[, 4])))
gi_PR <- GInteractions(gr1, gr2, mode = "strict")
# Construct name
gi_PR$name <- paste("Width", calculateDistances(gi_PR), sep = ":")
# Load CR data
mtx_CR <- read.table(fileNameIn2, sep = "\t", header = FALSE)
gr1 <- data.frame(chr = mtx_CR[, 1],
start = mtx_CR[, 2],
end = mtx_CR[, 3]) %>% makeGRangesFromDataFrame(., seqinfo = unique(c(mtx_CR[, 1], mtx_CR[, 4])))
gr2 <- data.frame(chr = mtx_CR[, 4],
start = mtx_CR[, 5],
end = mtx_CR[, 6]) %>% makeGRangesFromDataFrame(., seqinfo = unique(c(mtx_CR[, 1], mtx_CR[, 4])))
gi_CR <- GInteractions(gr1, gr2, mode = "strict")
# Construct name
gi_CR$name <- paste("Width", calculateDistances(gi_CR), sep = ":")
# Make seqlevels the same
seqlevels(gi_PR) <- seqlevels(gi_CR) <- unique(c(seqlevels(gi_PR), seqlevels(gi_CR)))
```
## SpectralTAD
```{r spectraltad, eval=FALSE}
# Neoloopfinder data
dir_data <- "/Users/mdozmorov/Documents/Data/GoogleDrive/HiC_files/results/TAD_Boundaries/SpectralTAD_v2/"
fileNameIn1 <- file.path(dir_data, paste0("UCD52PR_VC_10000_qualT_zT_", res, "_0.8.pairedbed")) # PR
fileNameIn2 <- file.path(dir_data, paste0("UCD52CR_VC_10000_qualT_zT_", res, "_0.8.pairedbed")) # CR
# Function to construct GInteractions
# Should be modified for different data
mtx2GInteractions <- function(fileNameIn = fileNameIn1, condition = "PR") {
# Load data
mtx <- read.table(fileNameIn, sep = "\t", header = FALSE)
gr1 <- data.frame(chr = mtx[, 1],
start = mtx[, 2],
end = mtx[, 3]) %>% makeGRangesFromDataFrame(., seqinfo = unique(c(mtx[, 1], mtx[, 4])))
gr2 <- data.frame(chr = mtx[, 4],
start = mtx[, 5],
end = mtx[, 6]) %>% makeGRangesFromDataFrame(., seqinfo = unique(c(mtx[, 1], mtx[, 4])))
# Domain GRanges overlapping centromeres
# gr <- data.frame(chr = mtx[, 1],
# start = mtx[, 2],
# end = mtx[, 6]) %>% makeGRangesFromDataFrame(., seqinfo = unique(mtx[, 1]))
gr1_centromere <- findOverlaps(gr1, hg38.UCSC.centromere, type = "any")
gr2_centromere <- findOverlaps(gr2, hg38.UCSC.centromere, type = "any")
# Boundaries overlapping telomeres
gr1_telomeres <- findOverlaps(gr1, hg38.UCSC.telomere, type = "any")
gr2_telomeres <- findOverlaps(gr2, hg38.UCSC.telomere, type = "any")
# Boundaries overlapping excludable regions
gr1_exclude <- findOverlaps(gr1, excludeGR.hg38.Kundaje.1, type = "any")
gr2_exclude <- findOverlaps(gr2, excludeGR.hg38.Kundaje.1, type = "any")
# Index to exclude
index_exclude <- sort(unique(c(queryHits(gr1_centromere), queryHits(gr2_centromere), queryHits(gr1_telomeres), queryHits(gr2_telomeres), queryHits(gr1_exclude), queryHits(gr2_exclude))))
# Complement index to include
index_include <- setdiff(seq(1:length(gr1)), index_exclude)
# GInteractions excluding selected boundaries
gi <- GInteractions(gr1[index_include], gr2[index_include], mode = "strict")
# Construct name
gi$name <- paste(condition, paste("Width", calculateDistances(gi), sep = ":"), sep = "_")
return(gi)
}
# Create condition-specific GInteractions
gi_PR <- mtx2GInteractions(fileNameIn = fileNameIn1, condition = "PR")
gi_CR <- mtx2GInteractions(fileNameIn = fileNameIn2, condition = "CR")
# Make seqlevels the same
seqlevels(gi_PR) <- seqlevels(gi_CR) <- unique(c(seqlevels(gi_PR), seqlevels(gi_CR)))
```
## hicFindTADs
```{r hicFindTADs, eval=FALSE}
# hicFindTADs data
dir_data <- "/Users/mdozmorov/Documents/Data/GoogleDrive/HiC_files/results/TAD_Boundaries/hicFindTADs_v2"
fileNameIn1 <- file.path(dir_data, "UCD52PR_10000_norm_KR_0.0005_domains.bed")
fileNameIn2 <- file.path(dir_data, "UCD52CR_10000_norm_KR_0.0005_domains.bed")
# Function to construct GInteractions
# Should be modified for different data
mtx2GInteractions <- function(fileNameIn = fileNameIn1, condition = "PR") {
# Load data
mtx <- read.table(fileNameIn, sep = "\t", header = FALSE)
gr1 <- data.frame(chr = mtx[, 1],
start = mtx[, 2] - (as.numeric(res) * 1000 / 2),
end = mtx[, 2] + (as.numeric(res) * 1000 / 2)) %>% makeGRangesFromDataFrame(., seqinfo = unique(mtx[, 1]))
gr2 <- data.frame(chr = mtx[, 1],
start = mtx[, 3] - (as.numeric(res) * 1000 / 2),
end = mtx[, 3] + (as.numeric(res) * 1000 / 2)) %>% makeGRangesFromDataFrame(., seqinfo = unique(mtx[, 1]))
# Domain GRanges overlapping centromeres
# gr <- data.frame(chr = mtx[, 1],
# start = mtx[, 2],
# end = mtx[, 6]) %>% makeGRangesFromDataFrame(., seqinfo = unique(mtx[, 1]))
gr1_centromere <- findOverlaps(gr1, hg38.UCSC.centromere, type = "any")
gr2_centromere <- findOverlaps(gr2, hg38.UCSC.centromere, type = "any")
# Boundaries overlapping telomeres
gr1_telomeres <- findOverlaps(gr1, hg38.UCSC.telomere, type = "any")
gr2_telomeres <- findOverlaps(gr2, hg38.UCSC.telomere, type = "any")
# Boundaries overlapping excludable regions
gr1_exclude <- findOverlaps(gr1, excludeGR.hg38.Kundaje.1, type = "any")
gr2_exclude <- findOverlaps(gr2, excludeGR.hg38.Kundaje.1, type = "any")
# Index to exclude
index_exclude <- sort(unique(c(queryHits(gr1_centromere), queryHits(gr2_centromere), queryHits(gr1_telomeres), queryHits(gr2_telomeres), queryHits(gr1_exclude), queryHits(gr2_exclude))))
# Complement index to include
index_include <- setdiff(seq(1:length(gr1)), index_exclude)
# GInteractions excluding selected boundaries
gi <- GInteractions(gr1[index_include], gr2[index_include], mode = "strict")
# Construct name
gi$name <- paste(condition, paste("Width", calculateDistances(gi), sep = ":"), mtx[index_include, 4], round(mtx[index_include, 5], digits = 2), sep = "_")
return(gi)
}
# Create condition-specific GInteractions
gi_PR <- mtx2GInteractions(fileNameIn = fileNameIn1, condition = "PR")
gi_CR <- mtx2GInteractions(fileNameIn = fileNameIn2, condition = "CR")
# Make seqlevels the same
seqlevels(gi_PR) <- seqlevels(gi_CR) <- unique(c(seqlevels(gi_PR), seqlevels(gi_CR)))
```
# Results file names
```{r results}
# Results
dir_results <- file.path(dir_data, paste0("preprocessing_", overlap_type))
dir.create(dir_results, recursive = TRUE) # Create if does not exist
## Loops
fileNameOut1.1 <- file.path(dir_results, paste0("loops_PR_unique_", res, "kb.bedpe"))
fileNameOut1.2 <- file.path(dir_results, paste0("loops_CR_unique_", res, "kb.bedpe"))
fileNameOut1.3 <- file.path(dir_results, paste0("loops_PR_common_", res, "kb.bedpe"))
fileNameOut1.4 <- file.path(dir_results, paste0("loops_CR_common_", res, "kb.bedpe"))
fileNameOut1.5 <- file.path(dir_results, paste0("loops_combined_", res, "kb.bedpe"))
fileNameOut1.6 <- file.path(dir_results, paste0("loops_PR_all_", res, "kb.bedpe"))
fileNameOut1.7 <- file.path(dir_results, paste0("loops_CR_all_", res, "kb.bedpe"))
## Anchors
fileNameOut2.1 <- file.path(dir_results, paste0("anchors_PR_unique_", res, "kb.bed"))
fileNameOut2.2 <- file.path(dir_results, paste0("anchors_CR_unique_", res, "kb.bed"))
fileNameOut2.3 <- file.path(dir_results, paste0("anchors_PR_common_", res, "kb.bed"))
fileNameOut2.4 <- file.path(dir_results, paste0("anchors_CR_common_", res, "kb.bed"))
fileNameOut2.5 <- file.path(dir_results, paste0("anchors_combined_", res, "kb.bed"))
fileNameOut2.6 <- file.path(dir_results, paste0("anchors_PR_all_", res, "kb.bed"))
fileNameOut2.7 <- file.path(dir_results, paste0("anchors_CR_all_", res, "kb.bed"))
# Statistics with the number of loops and anchors
fileNameOut3 <- file.path(dir_results, paste0("log_", res, "kb.csv"))
```
# Separate loops
```{r separate_loops}
# Intersect loops
index_common_loops <- findOverlaps(gi_PR, gi_CR, type = overlap_type)
# Condition-specific loops complementing common
loops_PR_unique <- gi_PR[setdiff(1:length(gi_PR), unique(queryHits(index_common_loops)))]
loops_CR_unique <- gi_CR[setdiff(1:length(gi_CR), unique(subjectHits(index_common_loops)))]
# Common loops should have identical first and second anchors. Don't check if no overlap, length(index_common_loops) is 0
if (length(index_common_loops) != 0) {
# Overlap between first anchors
overlap_common_loops1 <- findOverlaps(anchorOne(gi_PR[queryHits(index_common_loops)]), anchorOne(gi_CR[subjectHits(index_common_loops)]), type=overlap_type)
# Overlap between second anchors
overlap_common_loops2 <- findOverlaps(anchorTwo(gi_PR[queryHits(index_common_loops)]), anchorTwo(gi_CR[subjectHits(index_common_loops)]), type=overlap_type)
# Their indices should be equal
stopifnot(all.equal(unique(sort(queryHits(overlap_common_loops1))),
unique(sort(subjectHits(overlap_common_loops1)))) &
all.equal(unique(sort(queryHits(overlap_common_loops2))),
unique(sort(subjectHits(overlap_common_loops2)))))
# stopifnot(all.equal(anchorOne(gi_PR[queryHits(index_common_loops)]), anchorOne(gi_CR[subjectHits(index_common_loops)])) &
# all.equal(anchorTwo(gi_PR[queryHits(index_common_loops)]), anchorTwo(gi_CR[subjectHits(index_common_loops)])))
}
# Actual common loops
loops_PR_common <- gi_PR[unique(queryHits(index_common_loops))]
loops_CR_common <- gi_CR[unique(subjectHits(index_common_loops))]
# PR unique - green, 0,255,0
bedpe_PR_unique <- toBEDPE(mtx_selected = as.data.frame(loops_PR_unique), col = "0,255,0", condition = "PR")
fwrite(bedpe_PR_unique, file = fileNameOut1.1, quote = FALSE, sep = "\t", row.names = FALSE)
# CR unique - red, 255,0,0
bedpe_CR_unique <- toBEDPE(mtx_selected = as.data.frame(loops_CR_unique), col = "255,0,0", condition = "CR")
fwrite(bedpe_CR_unique, file = fileNameOut1.2, quote = FALSE, sep = "\t", row.names = FALSE)
# Common - blue, 0,0,255
# If common are present, add them to the combined, otherwise, combine just PR and CR specific
# if (length(index_common_loops) != 0) {
# bedpe_XX_common <- toBEDPE(mtx_selected = as.data.frame(loops_XX_common), col = "0,0,255", condition = "Common")
# fwrite(bedpe_XX_common, file = fileNameOut1.3, quote = FALSE, sep = "\t", row.names = FALSE)
# # Save combined, with common
# fwrite(rbind(bedpe_PR_unique, bedpe_CR_unique, bedpe_XX_common), file = fileNameOut1.5, quote = FALSE, sep = "\t", row.names = FALSE)
# } else {
# # Save combined, without common
# fwrite(rbind(bedpe_PR_unique, bedpe_CR_unique), file = fileNameOut1.5, quote = FALSE, sep = "\t", row.names = FALSE)
# }
if (length(index_common_loops) != 0) {
# PR common
bedpe_PR_common <- toBEDPE(mtx_selected = as.data.frame(loops_PR_common), col = "0,0,255", condition = "Common PR")
fwrite(bedpe_PR_common, file = fileNameOut1.3, quote = FALSE, sep = "\t", row.names = FALSE)
# CR common
bedpe_CR_common <- toBEDPE(mtx_selected = as.data.frame(loops_CR_common), col = "0,0,255", condition = "Common CR")
fwrite(bedpe_CR_common, file = fileNameOut1.4, quote = FALSE, sep = "\t", row.names = FALSE)
# Save combined, with common
fwrite(rbind(bedpe_PR_unique, bedpe_CR_unique, bedpe_PR_common, bedpe_CR_common), file = fileNameOut1.5, quote = FALSE, sep = "\t", row.names = FALSE)
} else {
# Save combined, without common
fwrite(rbind(bedpe_PR_unique, bedpe_CR_unique), file = fileNameOut1.5, quote = FALSE, sep = "\t", row.names = FALSE)
}
# Save statistics
fwrite(list("Loops PR all count", length(gi_PR)), file = fileNameOut3, append = TRUE)
fwrite(list("Loops CR all count", length(gi_CR)), file = fileNameOut3, append = TRUE)
fwrite(list("Loops PR unique count", length(loops_PR_unique)), file = fileNameOut3, append = TRUE)
fwrite(list("Loops CR unique count", length(loops_CR_unique)), file = fileNameOut3, append = TRUE)
if (length(index_common_loops) != 0) {
fwrite(list("Loops PR common count", length(loops_PR_common)), file = fileNameOut3, append = TRUE)
fwrite(list("Loops CR common count", length(loops_CR_common)), file = fileNameOut3, append = TRUE)
}
# Width summary
fwrite(width_summary("Loops PR all width", gi_PR), file = fileNameOut3, append = TRUE)
fwrite(width_summary("Loops CR all width", gi_CR), file = fileNameOut3, append = TRUE)
fwrite(width_summary("Loops PR unique width", loops_PR_unique), file = fileNameOut3, append = TRUE)
fwrite(width_summary("Loops CR unique width", loops_CR_unique), file = fileNameOut3, append = TRUE)
if (length(index_common_loops) != 0) {
fwrite(width_summary("Loops PR common width", loops_PR_common), file = fileNameOut3, append = TRUE)
fwrite(width_summary("Loops CR common width", loops_CR_common), file = fileNameOut3, append = TRUE)
}
# Difference analysis
fwrite(list("PR all vs. CR all width difference", width_difference(gi_PR, gi_CR)), file = fileNameOut3, append = TRUE)
fwrite(list("PR unique vs. CR unique width difference", width_difference(loops_PR_unique, loops_CR_unique)), file = fileNameOut3, append = TRUE)
if (length(index_common_loops) != 0) {
fwrite(list("PR unique vs. PR common width difference", width_difference(loops_PR_unique, loops_PR_common)), file = fileNameOut3, append = TRUE)
fwrite(list("CR unique vs. CR common width difference", width_difference(loops_CR_unique, loops_CR_common)), file = fileNameOut3, append = TRUE)
}
```
# Separate anchors
BED files are sorted by anchor interaction frequency, most frequently interacting on top
```{r separate_anchors}
## Get unique anchors with interaction frequency counts
# PR
# Combine first and second anchors
all_anchors_PR <- c(anchors(gi_PR)$first, anchors(gi_PR)$second) %>% sort()
# Count overlap with themselves = frequency
all_anchors_PR$freq <- countOverlaps(all_anchors_PR, all_anchors_PR, type = overlap_type)
full_anchors_PR <- all_anchors_PR # Unreduced anchors
# Reduce adjacent anchors
all_anchors_PR <- reduce(all_anchors_PR)
# CR
# Combine first and second anchors
all_anchors_CR <- c(anchors(gi_CR)$first, anchors(gi_CR)$second) %>% sort()
# Count overlap with themselves = frequency
all_anchors_CR$freq <- countOverlaps(all_anchors_CR, all_anchors_CR, type = overlap_type)
full_anchors_CR <- all_anchors_CR # Unreduced anchors
# Reduce adjacent anchors
all_anchors_CR <- reduce(all_anchors_CR)
# Selected region, for testing
# gr_selected <- GRanges(seqnames="chr7", ranges = IRanges(20455284, 20737829))
# all_anchors_PR <- subsetByOverlaps(all_anchors_PR, gr_selected)
# all_anchors_CR <- subsetByOverlaps(all_anchors_CR, gr_selected)
## Indexes for overlapping regions
index_common_anchors <- findOverlaps(all_anchors_PR, all_anchors_CR, type = overlap_type)
## Indexes for unique regions
index_anchors_PR_unique <- setdiff(1:length(all_anchors_PR), queryHits(index_common_anchors))
index_anchors_CR_unique <- setdiff(1:length(all_anchors_CR), subjectHits(index_common_anchors))
# PR
anchors_PR_unique <- all_anchors_PR[index_anchors_PR_unique] # All unique anchors
# anchors_PR_unique <- anchors_PR_unique[order(anchors_PR_unique$freq, decreasing = TRUE)] # Most interacting on top
# CR
anchors_CR_unique <- all_anchors_CR[index_anchors_CR_unique] # All unique anchors
# anchors_CR_unique <- anchors_CR_unique[order(anchors_CR_unique$freq, decreasing = TRUE)] # Most interacting on top
# All common anchors
anchors_PR_common <- all_anchors_PR[unique(queryHits(index_common_anchors))]
anchors_CR_common <- all_anchors_CR[unique(subjectHits(index_common_anchors))]
# Restore the original anchor coordinates
anchors_PR_unique <- subsetByOverlaps(full_anchors_PR, anchors_PR_unique)
anchors_CR_unique <- subsetByOverlaps(full_anchors_CR, anchors_CR_unique)
anchors_PR_common <- subsetByOverlaps(full_anchors_PR, anchors_PR_common)
anchors_CR_common <- subsetByOverlaps(full_anchors_CR, anchors_CR_common)
# Check, should be equal reduced. Don't check if no overlap, length(index_common_loops) is 0
# if (length(index_common_anchors) != 0) {
# stopifnot(all.equal(reduce(anchors_PR_common), reduce(anchors_CR_common)))
# }
# # Check, common regions should be equal. Don't check if no overlap, length(index_common_loops) is 0
# if (length(index_common_anchors) != 0) {
# # Overlap between reduced common regions should have the same indices
# overlap_common_anchors <- findOverlaps(reduce(anchors_PR_common),
# reduce(anchors_CR_common),
# type = overlap_type)
# stopifnot(all.equal(unique(sort(queryHits(overlap_common_anchors))), unique(sort(subjectHits(overlap_common_anchors)))))
#
# # stopifnot(all.equal(granges(all_anchors_PR[queryHits(index_common_anchors)]), granges(all_anchors_CR[subjectHits(index_common_anchors)])))
# }
# PR unique - green, 0,255,0
bed_PR_unique <- toBED(mtx_selected = as.data.frame(anchors_PR_unique), col = "0,255,0", condition = "PR", header = TRUE)
fwrite(bed_PR_unique, file = fileNameOut2.1, quote = FALSE, sep = "\t", row.names = FALSE, col.names = FALSE)
# CR unique - red, 255,0,0
bed_CR_unique <- toBED(mtx_selected = as.data.frame(anchors_CR_unique), col = "255,0,0", condition = "CR", header = TRUE)
fwrite(bed_CR_unique, file = fileNameOut2.2, quote = FALSE, sep = "\t", row.names = FALSE, col.names = FALSE)
# Common - blue, 0,0,255
# If common are present, add them to the combined, otherwise, combine just PR and CR specific
if (length(index_common_anchors) != 0) {
# PR common
bed_PR_common <- toBED(mtx_selected = as.data.frame(anchors_PR_common), col = "0,0,255", condition = "Common PR", header = TRUE)
fwrite(bed_PR_common, file = fileNameOut2.3, quote = FALSE, sep = "\t", row.names = FALSE, col.names = FALSE)
# CR common
bed_CR_common <- toBED(mtx_selected = as.data.frame(anchors_CR_common), col = "0,0,255", condition = "Common CR", header = TRUE)
fwrite(bed_CR_common, file = fileNameOut2.4, quote = FALSE, sep = "\t", row.names = FALSE, col.names = FALSE)
# Create combined, with common
bed_combined <- rbind(toBED(mtx_selected = as.data.frame(anchors_PR_unique), col = "0,255,0", condition = "PR", header = FALSE),
toBED(mtx_selected = as.data.frame(anchors_CR_unique), col = "255,0,0", condition = "CR", header = FALSE),
toBED(mtx_selected = as.data.frame(anchors_PR_common), col = "0,0,255", condition = "Common PR", header = FALSE) ,
toBED(mtx_selected = as.data.frame(anchors_CR_common), col = "0,0,255", condition = "Common CR", header = FALSE))
# Sort by frequency
bed_combined <- bed_combined[order(bed_combined$score, decreasing = TRUE), ]
# Append header
bed_combined <- rbind(c("track itemRgb=On", rep("", ncol(bed_combined) - 1)), bed_combined)
# Save combined
fwrite(bed_combined, file = fileNameOut2.5, quote = FALSE, sep = "\t", row.names = FALSE, col.names = FALSE)
} else {
# Create combined, without common
bed_combined <- rbind(toBED(mtx_selected = as.data.frame(anchors_PR_unique), col = "0,255,0", condition = "PR", header = FALSE),
toBED(mtx_selected = as.data.frame(anchors_CR_unique), col = "255,0,0", condition = "CR", header = FALSE))
# Sort by frequency
bed_combined <- bed_combined[order(bed_combined$score, decreasing = TRUE), ]
# Append header
bed_combined <- rbind(c("track itemRgb=On", rep("", ncol(bed_combined) - 1)), bed_combined)
# Save combined
fwrite(bed_combined, file = fileNameOut2.5, quote = FALSE, sep = "\t", row.names = FALSE, col.names = FALSE)
}
# Save statistics
fwrite(list("Anchors PR all count", length(full_anchors_PR)), file = fileNameOut3, append = TRUE)
fwrite(list("Anchors CR all count", length(full_anchors_CR)), file = fileNameOut3, append = TRUE)
fwrite(list("Anchors PR unique count", length(anchors_PR_unique)), file = fileNameOut3, append = TRUE)
fwrite(list("Anchors CR unique count", length(anchors_CR_unique)), file = fileNameOut3, append = TRUE)
fwrite(list("Anchors PR common count", length(anchors_PR_common)), file = fileNameOut3, append = TRUE)
fwrite(list("Anchors CR common count", length(anchors_CR_common)), file = fileNameOut3, append = TRUE)
```
# All loops
```{r all_loops}
# PR all - green, 0,255,0
bedpe_PR_all <- toBEDPE(mtx_selected = as.data.frame(gi_PR), col = "0,255,0", condition = "PR all")
fwrite(bedpe_PR_all, file = fileNameOut1.6, quote = FALSE, sep = "\t", row.names = FALSE)
# CR all - red, 255,0,0
bedpe_CR_all <- toBEDPE(mtx_selected = as.data.frame(gi_CR), col = "255,0,0", condition = "CR all")
fwrite(bedpe_CR_all, file = fileNameOut1.7, quote = FALSE, sep = "\t", row.names = FALSE)
```
# All anchors
```{r all_anchors}
# PR all - green, 0,255,0
bed_PR_all <- toBED(mtx_selected = as.data.frame(full_anchors_PR), col = "0,255,0", condition = "PR", header = TRUE)
fwrite(bed_PR_all, file = fileNameOut2.6, quote = FALSE, sep = "\t", row.names = FALSE, col.names = FALSE)
# CR all - red, 255,0,0
bed_CR_all <- toBED(mtx_selected = as.data.frame(full_anchors_CR), col = "255,0,0", condition = "CR", header = TRUE)
fwrite(bed_CR_all, file = fileNameOut2.7, quote = FALSE, sep = "\t", row.names = FALSE, col.names = FALSE)
```