環狀基因表達熱圖.png

1.讀取基因表達數據

這是我的基因表達量數據

Fig1.PNG

myfiles <- list.files(pattern = "*.csv")
myfiles
matrix<-read.csv(myfiles[1],sep=',',header=T,check.names = FALSE,row.names = 1)

1.1提取部分數據集

matrix<-subset(matrix,padj<0.2)

1.2提取基因表達值所在的列，組成新的矩陣，并將矩陣轉置

為什么要轉置，參考http://www.lxweimin.com/p/115d07af3029

mat =t(matrix[,7:12])

7-12列為每個樣本的基因表達量

1.3基因表達歸一化

mat=scale(mat, center = TRUE, scale = TRUE)
View(mat)

scale函數可以對矩陣歸一化；參考http://www.lxweimin.com/p/115d07af3029

1.4對數據進行聚類,從而得到其dendrogram對象

dend <-as.dendrogram(hclust(dist(t(mat))))

dist函數計算microRNA間的距離，hclust來進行層次聚類

1.5定義進化樹顏色

library(dendextend)
n=3
dend <-dend %>% set("branches_k_color", k = n) 

n可自定義

1.6可視化

par(mar=c(7.5,3,1,0))
plot(dend)

Fig2.png

1.7聚類后的矩陣

如圖Fig 2，聚類后的矩陣的列的順序會發生變化。按此順序，重新排列矩陣。

mat2 = mat[, order.dendrogram(dend)]

1.7.1查看矩陣重排后的樣本名

lable1=row.names(mat2);lable1
[1] "H-CK-1-3" "H-CK-2-3" "H-CK-3-3" "H-PA-1-3" "H-PA-2-3" "H-PA-3-3"

1.7.2查看矩陣重排后的基因名

lable2=colnames(mat2);lable2

只有基因名順序，也就是列名順序發生變化。

nr = nrow(mat2);nr
[1] 6
nc = ncol(mat2);nc
[1] 86

1.8 構建顏色轉變函數

require("circlize")
col_fun = colorRamp2(c(-1.5, 0, 1.5), c("skyblue", "white", "red"))

1.9 矩陣中的數值轉變為顏色

 col_mat = col_fun(mat2)

1.9.1 查看第1列

col_mat[,1]

1.9.1.1 結果

 H-CK-1-3    H-CK-2-3    H-CK-3-3    H-PA-1-3    H-PA-2-3    H-PA-3-3 
"#FF0000FF" "#FFDED3FF" "#FFAF96FF" "#ABDBF1FF" "#DCF0F9FF" "#BDE3F4FF" 

2.1 畫板初始化設置

par(mar=c(0,0,0,0))
circos.clear();circos.par(canvas.xlim =c(-1.3,1.3),
                          canvas.ylim = c(-1.3,1.3),
                          cell.padding = c(0,0,0,0), 
                          gap.degree =90)
factors = "a"
circos.initialize(factors, xlim = c(0, ncol(mat2)))

xlim和ylim設置1.3，是為了防止添加基因名時溢出

2.2 添加第一個軌道

circos.track(ylim = c(0, nr),bg.border = NA,track.height = 0.1*nr,
             panel.fun = function(x, y) {
               for(i in 1:nr) {
                 circos.rect(xleft = 1:nc - 1, ybottom = rep(nr - i, nc),
                             xright = 1:nc, ytop = rep(nr - i + 1, nc),
                             border = "white",
                             col = col_mat[i,])
                 circos.text(x = nc,
                             y = 6.4 -i,
                             labels = lable1[i],
                             facing = "downward", niceFacing = TRUE,
                             cex = 0.6,
                             adj = c(-0.2, 0))
                 }
})

2.3 添加基因名稱

for(i in 1:nc){
  circos.text(x = i-0.4,
              y = 7,
              labels = lable2[i],
              facing = "clockwise", niceFacing = TRUE,
              cex = 0.4,adj = c(0, 0))
  }

2.4 添加進化樹

max_height <-max(attr(dend, "height"))
circos.track(ylim = c(0, max_height),bg.border = NA,track.height = 0.3, 
             panel.fun = function(x, y){
               circos.dendrogram(dend = dend,
                                 max_height = max_height)
             })
circos.clear()

2.5 添加圖例

library(ComplexHeatmap)
lgd <- Legend(at = c(-2,-1, 0, 1, 2), col_fun = col_fun, 
              title_position = "topcenter",title = "Z-score")
draw(lgd, x = unit(0.7, "npc"), y = unit(0.7, "npc"))

參考資料：
https://www.shenxt.info/post/r-circlize-study-2/