BioScripts/locuszoom.R at main · Zilong-Li/BioScripts · GitHub

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
#!/usr/bin/env Rscript


locusZoomNoLD <- function(pval, pos, chr, refGenes = NULL, w = 1, plog = FALSE, main = "LocusZoom") {
  xforrs <- 0.03
  regsz <- 1.2
  width <- 22
  rsqplus <- 0.045
  rightylabxplus <- 0.05
  xmargin <- 0.005
  cexaxis <- 1.3
  cexlab <- 1.5
  blue <- "dodgerblue4"

  pval[is.na(pval)] <- 1

  N <- length(pval)

  rsqwithn <- rep(0, N - 1)
  ## palette(brewer.pal(10, "RdBu"))[1:10]
  cols <- c(10, 8, 6, 4, 2, 2)
  getcol <- function(x) {
    cuts <- c(.2, .4, .6, .8, 1)
    cols[6 - sum(as.numeric(x) < cuts)]
  }

  min.pos <- min(pos) / 1e6
  max.pos <- max(pos) / 1e6

  adj <- 0
  nf <- layout(matrix(c(1, 2), 2, 1, byrow = TRUE), heights = c(4.0, 2.5), widths = c(9))
  par(mar = c(0, 5.5, 2, 1))

  bgcols <- sapply(rsqwithn, getcol)
  bgcols[w] <- 1
  if (plog) {
    ylim <- c(0, max(10.5, max(-log10(pval))))
    pval <- -log10(pval)
  } else {
    ylim <- c(0, max(pval)*1.05)
  }
  plot(pos / 1e6, pval, col = "black", bg = bgcols, xaxs = "i", xlim = c(min.pos - adj, max.pos + adj), xaxt = "n", xlab = "", ylab = "", ylim = ylim, pch = 21, cex = 2, las = 1, cex.axis = cexaxis, main = main)
  ## axis(2, ylim = c(0, 15), col = "black", label = FALSE, cex.axis = cexaxis)
  ## mtext(2, text = expression(paste("-l", og[10], "[", italic(P), "]")), line = 3, cex = cexlab)
  mtext(2, text = "SNP Loadings of PC1", line = 4, cex = cexlab)
  xstart <- max.pos - (max.pos - min.pos) * 0.15
  scalewidth <- (max.pos - min.pos) * 0.0125
  xx <- c(xstart, xstart, xstart + scalewidth, xstart + scalewidth)
  #  legend("topleft",legend=c(rs),pt.bg=1,col="black",cex=1.6,pch=21)

  ybot <- ylim[2] * 0.65
  ytop <- ylim[2] * 0.95
  ysz <- ytop - ybot
  cuts <- c(.2, .4, .6, .8, 1)
  txtcuts <- c("0.2", "0.4", "0.6", "0.8", "1.0")
  txtposs <- c(0.2, 0.4, 0.6, 0.8, 1.00)
  scalexplus <- (max.pos - min.pos) * 0.06
  ## if (FALSE) {
  ##   polygon(xx + scalexplus, c(ybot, ybot + cuts[1] * ysz, ybot + cuts[1] * ysz, ybot), col = cols[1])

  ##   for (i in 2:length(cols)) {
  ##     polygon(xx + scalexplus, c(ybot + cuts[(i - 1)] * ysz, ybot + cuts[i] * ysz, ybot + cuts[i] * ysz, ybot + ysz * cuts[(i - 1)]), col = cols[i])
  ##   }
  ##   scalenumplus <- (max.pos - min.pos) * 0.09
  ##   # text((xx[1]+0.4+xx[3])/2+rsqplus,ytop+0.35,expression(italic(r)^2),cex=cexaxis)
  ##   text(xx[3] + scalenumplus, ytop + 0.35, expression(italic(r)^2), cex = cexaxis)

  ##   text(rep(xx[3] + scalenumplus, 2), txtposs[3:4] * ysz + ybot - 0.07, txtcuts[3:4], cex = cexlab - 0.1)
  ##   text(rep(xx[3] + scalenumplus, 1), txtposs[2] * ysz + ybot - 0.07, txtcuts[2], cex = cexlab - 0.1)
  ##   text(rep(xx[3] + scalenumplus, 1), txtposs[1] * ysz + ybot - 0.07, txtcuts[1], cex = cexlab - 0.1)
  ## }

  dat <- read.table(refGenes, as.is = T, head = T, comment.char = "")
  xx2 <- dat[dat[, "chrom"] == paste("chr", chr, sep = "") & dat[, "cdsStart"] < max.pos * 1e6 & dat[, "cdsEnd"] > min.pos * 1e6, ]

  start <- xx2$txStart # casia. this column is needed in the refgene file
  end <- xx2$txEnd # casia. this column is needed in the refgene file
  nams <- xx2$name2 # casia. this column is needed in the refgene file
  cnts <- xx2$exonCount # casia. this column is needed in the refgene file but you can have all as NA

  abline(v = pos, lty = 2)
  ## par(mar = c(5.2, 6.2, -0.1, 6.3) + 0.1)
  par(mar = c(5, 5.5, 1, 1))

  plot(c(0, 0), c(0, 0), type = "n", xlim = c(min.pos - adj, max.pos + adj), ylim = c(-0.8, 0.1), xlab = "", xaxs = "i", yaxt = "n", ylab = "", main = "", cex.lab = 2.6, cex.axis = cexaxis, tck = -0.05)
  mtext(1, text = paste("Position on chromosome ", chr, " (Mb)", sep = ""), line = 3, cex = cexlab)
  abline(v = pos, lty = 2)

  ord <- order(start)
  start <- start[ord]
  end <- end[ord]
  exoncnts <- cnts[ord]
  nams <- nams[ord]
  keep <- !duplicated(nams)
  start <- start[keep]
  end <- end[keep]
  exoncnts <- cnts[keep]
  nams <- nams[keep]
  ord <- ord[keep]
  he <- rep(c(0, -0.18, -0.36, -0.54, -0.72), 100)[1:length(nams)] - 0.05
  if (length(start) > 0) {
    segments(start / 1e6, he, end / 1e6, he)
    keep <- !duplicated(nams)
    sapply(1:sum(keep), function(x) {
      text((end[keep][x] + start[keep][x]) / 2e6, he[keep][x] + 0.08, bquote(italic(.(nams[keep][x]))), cex = cexlab - 0.6)
    })
    estart <- as.numeric(unlist(sapply(xx2$exonStarts[ord], function(y) {
      strsplit(y, ",")[[1]]
    }))) / 1e6 # Casia outcomment
    eend <- as.numeric(unlist(sapply(xx2$exonEnds[ord], function(y) {
      strsplit(y, ",")[[1]]
    }))) / 1e6 # Casia outcomment
    rect(estart, rep(he, xx2$exonCount[ord]) - 0.01, eend, rep(he, xx2$exonCount[ord]) + 0.01, col = "black") # Casia outcomment
  }
}

refGenes <- "~/Downloads/refGene.hg38.txt.gz"
locusZoomNoLD(t$pc, t$pos, 5, refGenes = refGenes)


d <- read.table("t.loadings.chr5.pc1.csv", h = F)
colnames(d) <- c("chr", "pos", "pc")
t <- subset(d, pos > 33851693 & pos < 33991693)

head(sort(t$pc, de = T))

dev.off()

pos <- t$pos
min.pos <- min(pos) / 1e6
max.pos <- max(pos) / 1e6
head(xx2)