公众号“生信小课堂”

TCGA数据分析课程：生物信息学教学

相关性分析是我们生信分析中必不可少的技能，单基因的批量相关性分析，可以用于做单基因的GO，KEGG富集分析和GSEA分析，也有多基因之间的相关性分析，或者多个基因集之间的相关性分析。今天我们利用TCGA肝癌数据，挑选一些基因做相关性分析，并用不同的R包展示。

load(file = "mRNAdata.Rda")#加载数据

library(tidyverse)

library(corrplot)

library(circlize)TCGA数据下载与ID转换​mp.weixin.qq.com

mRNAdata

mRNAdata" />

选择列，我们随便选一些基因

new

new

计算相关性系数

cor_new

cor_new

corrplot包画图

method：指定形状，可以是circle圆形(默认)，square方形，ellipse,椭圆形，number数值，shade阴影，color颜色，pie饼图。type：指定显示范围，full完全(默认)，lower下三角，upper上三角

corrplot(cor_new, method = "circle")

corrplot(cor_new, method = "square")

corrplot(cor_new, method = "pie")

corrplot(cor_new, method = "color")

corrplot(cor_new, method = "number")

更改颜色

corrplot(cor_new, method = "color", col = colorRampPalette(c("blue", "#CD2626"))(10), title = "更改颜色")

corrplot(cor_new, method = "number", col = RColorBrewer::brewer.pal(n=8, name = "RdYlGn"))

聚类显示

corrplot(cor_new,method="color",order="hclust",title = "hclust聚类", diag = TRUE,hclust.method="average",addCoef.col = "blue")

组合展示

corrplot(cor_new, method = "circle", type = "upper", tl.pos = "d")

corrplot(cor_new, add = TRUE, type = "lower", method = "number", diag = FALSE, tl.pos = "n", cl.pos = "n")

做一个好看点的

col

corrplot(cor_new, method="color", col=col(200),

type="upper", order="hclust",

addCoef.col = "black", #添加相关系数

diag=FALSE

)

用circlize包展示

col_fun = colorRamp2(c(-1, 0, 1), c("#67BE54", "#FFFFFF", "#F82C2B"))

chordDiagram(cor_new, grid.col = 1:8, symmetric = TRUE, col = col_fun)

circos.clear()

circlize包作图非常好看，我们可以用它来做一个太极

#太极图

library(circlize)

factors = 1:8

circos.par(start.degree = 22.5, gap.degree = 6)

circos.initialize(factors = factors, xlim = c(0, 1))

# yang yao is __ (a long segment)

add_yang_yao = function() {

circos.rect(0,0,1,1, col = "black")

}

# yin yao is -- (two short segments)

add_yin_yao = function() {

circos.rect(0,0,0.45,1, col = "black")

circos.rect(0.55,0,1,1, col = "black")

}

circos.track(ylim = c(0, 1), factors = factors, bg.border = NA,

panel.fun = function(x, y) {

i = get.cell.meta.data("sector.numeric.index")

if(i %in% c(2, 5, 7, 8)) add_yang_yao() else add_yin_yao()

}, track.height = 0.1)

circos.track(ylim = c(0, 1), factors = factors, bg.border = NA,

panel.fun = function(x, y) {

i = get.cell.meta.data("sector.numeric.index")

if(i %in% c(1, 6, 7, 8)) add_yang_yao() else add_yin_yao()

}, track.height = 0.1)

circos.track(ylim = c(0, 1), factors = factors, bg.border = NA,

panel.fun = function(x, y) {

i = get.cell.meta.data("sector.numeric.index")

if(i %in% c(4, 5, 6, 7)) add_yang_yao() else add_yin_yao()

}, track.height = 0.1)

# the bottom of the most recent track

r = get.cell.meta.data("cell.bottom.radius") - 0.1

# draw taiji, note default order is clock wise for `draw.sector`

draw.sector(center = c(0, 0), start.degree = 90, end.degree = -90,

rou1 = r, col = "black", border = "black")

draw.sector(center = c(0, 0), start.degree = 270, end.degree = 90,

rou1 = r, col = "white", border = "black")

draw.sector(center = c(0, r/2), start.degree = 0, end.degree = 360,

rou1 = r/2, col = "white", border = "white")

draw.sector(center = c(0, -r/2), start.degree = 0, end.degree = 360,

rou1 = r/2, col = "black", border = "black")

draw.sector(center = c(0, r/2), start.degree = 0, end.degree = 360,

rou1 = r/8, col = "black", border = "black")

draw.sector(center = c(0, -r/2), start.degree = 0, end.degree = 360,

rou1 = r/8, col = "white", border = "white")

circos.clear()

飞镖盘

factors = 1:20 # just indicate there are 20 sectors

circos.par(gap.degree = 0, cell.padding = c(0, 0, 0, 0),

start.degree = 360/20/2, track.margin = c(0, 0), clock.wise = FALSE)

circos.initialize(factors = factors, xlim = c(0, 1))

circos.track(ylim = c(0, 1), factors = factors, bg.col = "black", track.height = 0.15)

circos.trackText(x = rep(0.5, 20), y = rep(0.5, 20),

labels = c(13, 4, 18, 1, 20, 5, 12, 9, 14, 11, 8, 16, 7, 19, 3, 17, 2, 15, 10, 6),

cex = 0.8, factors = factors, col = "#EEEEEE", font = 2, facing = "downward")

circos.track(ylim = c(0, 1), factors = factors,

bg.col = rep(c("#E41A1C", "#4DAF4A"), 10), bg.border = "#EEEEEE", track.height = 0.05)

circos.track(ylim = c(0, 1), factors = factors,

bg.col = rep(c("black", "white"), 10), bg.border = "#EEEEEE", track.height = 0.275)

circos.track(ylim = c(0, 1), factors = factors,

bg.col = rep(c("#E41A1C", "#4DAF4A"), 10), bg.border = "#EEEEEE", track.height = 0.05)

circos.track(ylim = c(0, 1), factors = factors,

bg.col = rep(c("black", "white"), 10), bg.border = "#EEEEEE", track.height = 0.375)

draw.sector(center = c(0, 0), start.degree = 0, end.degree = 360,

rou1 = 0.1, col = "#4DAF4A", border = "#EEEEEE")

draw.sector(center = c(0, 0), start.degree = 0, end.degree = 360,

rou1 = 0.05, col = "#E41A1C", border = "#EEEEEE")

circos.clear()

好了，多基因的相关性分析和作图就分享到这

公众号“生信小课堂”

TCGA数据分析课程：生物信息学教学