R语言Circlize包绘制和弦图

 library(circlize)
 mat <- matrix(1:9, 3) # 第1列不是id列，通过行命名替代
 rownames(mat) <-  letters[1:3]
 colnames(mat) <-  LETTERS[1:3]
 mat

 > set.seed(999)
 > bed <- generateRandomBed() # 默认参数
 > head(bed)
    chr   start     end     value1
 1 chr1   39485  159163 -0.1887635
 2 chr1  897195 1041959 -0.2435220
 3 chr1 1161957 1177159  0.3749953
 4 chr1 1201513 1481406 -0.2600839
 5 chr1 1487402 1531773 -0.4633990
 6 chr1 1769949 2736215 -0.8159909
 > 
 > bed <- generateRandomBed(nr = 200, nc = 4)
 > nrow(bed)
 [1] 205
 > 
 > bed <- generateRandomBed(nc = 2, fun = function(k) sample(letters, k, replace = TRUE)) # 默认产生1000行数据
 > head(bed)
    chr   start     end value1 value2
 1 chr1   98740  566688      e      e
 2 chr1  769960  887938      b      q
 3 chr1  906851  933021      u      o
 4 chr1 1241911 1243537      k      f
 5 chr1 1385344 1410947      v      x
 6 chr1 1498302 1585389      u      v

 > # 构造一个邻接矩阵
 > set.seed(999)
 > mat <- matrix(sample(18, 18), 3, 6) # 3行6列的矩阵
 > rownames(mat) <- paste0("S", 1:3) # 生成行名
 > colnames(mat) <- paste0("E", 1:6) # 生成列名
 > mat
    E1 E2 E3 E4 E5 E6
 S1  8 13 18  6 11 14
 S2 10 12  1  3  5  7
 S3  2 16  4 17  9 15
 > # 构造一个邻接列表
 > df <- data.frame(from = rep(rownames(mat), times = ncol(mat)), # 第1列对象
 +                  to = rep(colnames(mat), each = nrow(mat)), # 第2列对象
 +                  value = as.vector(mat),  # 第3列,前2列对象相互作用强度
 +                  stringsAsFactors = FALSE)
 > df
    from to value
 1    S1 E1     8
 2    S2 E1    10
 3    S3 E1     2
 4    S1 E2    13
 5    S2 E2    12
 6    S3 E2    16
 7    S1 E3    18
 8    S2 E3     1
 9    S3 E3     4
 10   S1 E4     6
 11   S2 E4     3
 12   S3 E4    17
 13   S1 E5    11
 14   S2 E5     5
 15   S3 E5     9
 16   S1 E6    14
 17   S2 E6     7
 18   S3 E6    15
 > 

 > library(circlize)
 > # 使用邻接矩阵
 > chordDiagram(mat)
 > circos.clear() # 结束绘图，否则会继续叠加图层
 > # 使用邻接列表
 > chordDiagram(df)
 > circos.clear()

 > library(circlize)
 > 
 > # 使用邻接矩阵时
 > circos.par(gap.after = c(rep(5, nrow(mat)-1),  # 2个5，表示3个行名之间的间隙分别为5个单位
 +                          15,                   # 表示行名与列名之间的间隙，为15个单位
 +                          rep(5, ncol(mat)-1),  # 5个5，表示6个列名之间的间隙分别为5个单位
 +                          15))                  # 表示列名与行名之间的间隙，为15个单位
 > chordDiagram(mat)
 > circos.clear() # 返回默认设置
 > 
 > # 使用邻接列表时
 > circos.par(gap.after = c(rep(5, length(unique(df[[1]]))-1), # 表示第1列元素之间的间隙为5个单位
 +                          15,                                # 表示第1列与第2列之间的间隙为15个单位
 +                          rep(5, length(unique(df[[2]]))-1), # 表示第2列元素之间的间隙为5个单位
 +                          15))                               # 表示第2列与第1列之间的间隙为15个单位 
 > chordDiagram(df)
 > circos.clear()

 > library(circlize)
 > 
 > circos.par(start.degree = 90, clock.wise = FALSE) # 逆时针旋转，起点位置在逆时针90度方向，即12点针方向 
 > chordDiagram(mat)
 > circos.clear()

 > library(circlize)
 > 
 > chordDiagram(mat, order = c("S1", "E1", "E2", "S2", "E3", "E4", "S3", "E5", "E6"))  # 使用order参数调整顺序，默认3点钟顺时针方向 
 > circos.clear()

 > library(circlize)
 > # 构建颜色向量，指定名称属性，则按名称匹配
 > grid_col <-  c(S1 = "red", S2 = "green", S3 = "blue",
 +     E1 = "grey", E2 = "grey", E3 = "grey", E4 = "grey", E5 = "grey", E6 = "grey") 
 ># 调整外围sector颜色，增加透明度
 > chordDiagram(mat, grid.col = grid_col, transparency = 0.7) 
 ># 按名称匹配，则link颜色与mat矩阵的列名一致，全变为灰色
 > chordDiagram(t(mat), grid.col = grid_col) 

 > 
 > # 数据是邻接矩阵
 > col_mat <- rand_color(length(mat), transparency = 0.7) # 产生随机颜色矩阵，并指定透明度
 > dim(col_mat) <- dim(mat) # 以确保col_mat是一个矩阵
 > chordDiagram(mat, grid.col = grid_col, col = col_mat) # 设置link颜色，
 > circos.clear()
 > 
 > # 数据是邻接列表
 > cols <- rand_color(nrow(df), transparency = 0.7)
 > chordDiagram(df, grid.col = grid_col, col = cols)
 > circos.clear()
 > 
 > # link为连续变量
 > col_fun <- colorRamp2(range(mat), c("#FFEEEE", "#FF0000"), transparency = 0.5) # 产生连续色块并指定透明度
 > chordDiagram(mat, grid.col = grid_col, col = col_fun)
 > circos.clear()
 > 
 > # 用数字指定link颜色
 > chordDiagram(mat, grid.col = grid_col, row.col = 1:3, transparency = 0.7) # 用数字向量指定颜色，向量长度与连接矩阵的行数相同
 > chordDiagram(mat, grid.col = grid_col, column.col = 1:6, transparency = 0.7) # 用数字向量指定颜色，向量长度与连接矩阵的列数相同
 > circos.clear()

 > library(circlize)
 > 
 > # 用长度的1的向量指定
 > chordDiagram(mat, grid.col = grid_col, link.lwd = 2, link.lty = 2, link.border = "red") # 指定link边线宽度，边线线型，边线颜色
 > circos.clear()
 > 
 > # 用矩阵指定
 > lwd_mat <- matrix(1, nrow = nrow(mat), ncol = ncol(mat)) # 元素为1的矩阵，其维度与源数据mat一致
 > lwd_mat[mat > 12] <- 2 # relation > 12,则加宽link边线
 > border_mat <- matrix(NA, nrow = nrow(mat), ncol = ncol(mat)) # 元素为NA的矩阵，其维度与源数据mat一致
 > border_mat[mat > 12] <- "red" # relation > 2，则为红色边缘线
 > chordDiagram(mat, grid.col = grid_col, link.lwd = lwd_mat, link.border = border_mat) # 指定link边缘线宽度，边缘线颜色
 > circos.clear()
 > 
 > # 参数矩阵维度与数据源不一致,则改变部分颜色,必须按名称属性匹配
 > border_mat2 <- matrix("black", nrow = 1, ncol = ncol(mat)) # 生成1行的矩阵，其宽与数据源mat一致
 > rownames(border_mat2) <- rownames(mat)[2] # 将mat第2个行名赋值给border_mat2，则只会改变第mat第2行的边缘线颜色
 > colnames(border_mat2) <- colnames(mat) # 赋值列名，与数据源mat一致
 > chordDiagram(mat, grid.col = grid_col, link.lwd = 2, link.border = border_mat2) #
 > circos.clear()
 > 
 > # 参数矩阵还可以设置为特殊的3列格式，前2列分别对应数据源的行名称和列名称，第3列为参数列，相当于邻接列表格式的参数矩阵
 > lty_df <- data.frame(c("S1", "S2", "S3"), c("E5", "E6", "E6"), c(1, 2, 3)) # link边缘线分别为1, 2, 3
 > lwd_df <- data.frame(c("S1", "S2", "S3"), c("E5", "E6", "E4"), c(2, 2, 2)) # link边线线宽为2
 > border_df <- data.frame(c("S1", "S2", "S3"), c("E5", "E6", "E4"), c(1, 1, 1)) # link边缘线颜色为1
 > chordDiagram(mat, grid.col = grid_col, link.lty = lty_df, link.lwd = lwd_df, link.border = border_df)
 > circos.clear()
 > 
 > # 当数据源是邻接列表时，只需要指定跟源数据一样行数的向量，特别方便
 > chordDiagram(df, grid.col = grid_col,
 +              link.lty = sample(1:3,nrow(df), replace = TRUE),
 +              link.lwd = runif(nrow(df)) * 2,
 +              link.border = sample(0:1, nrow(df), replace = TRUE))
 > circos.clear()

 > library(circlize)
 > 
 > # 通过指定不同的颜色进行高亮
 > chordDiagram(mat, grid.col = grid_col, row.col = c("#FF000080", "#00FF0010", "#0000FF10"))
 > circos.clear()
 > 
 > # 通过指定透明色给某些在范围之外的relation
 > col_mat[mat < 12] <- "#00000000" # relation < 12则变为透明色
 > chordDiagram(mat, grid.col = grid_col, col = col_mat) #
 > circos.clear()
 > 
 > # 通过函数同时指定透明色和高亮色，对邻接列表数据源也适用
 > col_fun <- function(x) {ifelse(x < 12, "#00000000", "#FF000080") }# relation小于12则为透明色，反之为#FF000080石榴红
 > chordDiagram(mat, grid.col = grid_col, col = col_fun, transparency = 0.7)
 > circos.clear()
 > 
 > # 事实上，所有颜色矩阵或颜色生成函数中色彩都是绘制在图形中的，只是程序内部将其透明度设置为了1，
 > # 通过3列特殊数据框指定的颜色，其缺失的颜色的relation将不会画出
 > col_df <- data.frame(c("S1","S2", "S3"), c("E5", "E6", "E4"),
 +                      c("#FF000080", "#00FF0080", "#0000FF80"))
 > chordDiagram(mat, grid.col = grid_col, col = col_df)
 > circos.clear()
 > 
 > # 对于邻接列表数据源，高亮弦调整更简单，只需要设置要高亮的颜色，其它为透明色就行了
 > cols <- rand_color(nrow(df))
 > cols[df[[3]] < 10] <- "#00000000" # 将df中第3列，即relation列，列值小于10的都更新为透明色HEX色值
 > chordDiagram(df, grid.col = grid_col, col = cols)
 > 
 > # 通过link.visible参数调整
 > cols <- rand_color(nrow(df))
 > chordDiagram(df, grid.col = grid_col, link.visible = df[[3]] >= 10) # 只显示df第3列大于10的弦
 > circos.clear()

 > library(circlize)
 > 
 > chordDiagram(mat, grid.col = grid_col, link.sort = TRUE, link.decreasing = TRUE) #按弦宽度下降排列,则弦宽沿顺时针方向逐渐下降
 > title("link.sort = TRUE, link.decreasing = TRUE",cex = 0.8) # 添加标题
 > circos.clear()
 > 
 > chordDiagram(mat, grid.col = grid_col, link.sort = TRUE, link.decreasing = FALSE) # 弦宽沿顺时针方向逐渐增大
 > title("link.sort = TRUE, link.decreasing = FALSE", cex = 0.8)
 > circos.clear()

 > library(circlize)
 > 
 > # 邻接矩阵数据源，求矩阵的秩，然后指定给link.rank参数
 > chordDiagram(mat, grid.col = grid_col, transparency = 0, link.rank = )# 设置透明度为0，方便观察
 > chordDiagram(mat, grid.col = grid_col, transparency = 0, link.rank = rank(mat)) # 用mat中的秩进行排序，秩最大先添加
 > circos.clear()
 > 
 > # 邻接列表数据源，对relation列求秩，然后指定给link.rank参数
 > chordDiagram(df, grid.col = grid_col, transparency = 0, link.rank = rank(df[[3]])) # 第3列为relation列，求秩
 > circos.clear()

 > library(circlize)
 > 
 > df2 <- data.frame(start = c("a", "b", "c", "a"), end = c("a", "a", "b", "c"))
 > chordDiagram(df2, grid.col = 1:3, self.link = 1) #
 > chordDiagram(df2, grid.col = 1:3, self.link = 2)
 > circos.clear()

 > library(circlize)
 > 
 > mat3 <- matrix(rnorm(25), 5) # 生成25个均匀分布的随机数, 5行排列
 > colnames(mat3) <- letters[1:5]
 > cor_mat <- cor(mat3) # 求相关系数,则变为对称矩阵
 > 
 > col_fun <- colorRamp2(c(-1, 0, 1), c("green", "white", "red"))
 > chordDiagram(cor_mat, grid.col = 1:5, symmetric = TRUE, col = col_fun)
 > title("symmetric = TRUE") # 增加标题
 > circos.clear()
 > 
 > chordDiagram(cor_mat, grid.col = 1:5, col = col_fun)
 > title("symmetric = FALSE")
 > circos.clear()

 > library(circlize)
 > 
 > par(mfrow = c(1, 3)) # 设置绘图环境，多图布局，1行3列布局
 > 
 > chordDiagram(mat, grid.col = grid_col, directional = 1) # 结束端要短一些
 > chordDiagram(mat, grid.col = grid_col, directional = 1, diffHeight = uh(5, "mm")) # 设定缩短量为5mm， uh表示传递单位 
 > chordDiagram(mat, grid.col = grid_col, directional = -1) # 反转方向，这行名对应的端要短一些
 > circos.clear()

 > library(circlize)
 > 
 > mat2 <- matrix(sample(100, 35), nrow = 5)
 > rownames(mat2) <- letters[1:5]
 > colnames(mat2) <- letters[1:7]
 > mat2
    a  b  c  d  e  f  g
 a 37 89  5 19 68 86 32
 b 23 71 18 75 62 34 16
 c 94 67 81 66 28 41 39
 d 72  9  4 12 17  6 44
 e 85 63  2 55 40 60 79
 > chordDiagram(mat2, grid.col = 1:7, directional = 1, row.col = 1:5)
 > circos.clear()

 > library(circlize)
 > 
 > mat3 <- mat2
 > for (cn in intersect(rownames(mat3), colnames(mat3))) {
 +   mat3[cn, cn] <- 0 # 将行名和列名相同的值更改为0
 + 
 + }
 > mat3
    a  b  c  d  e  f  g
 a  0 89  5 19 68 86 32
 b 23  0 18 75 62 34 16
 c 94 67  0 66 28 41 39
 d 72  9  4  0 17  6 44
 e 85 63  2 55  0 60 79
 > 
 > chordDiagram(mat3, grid.col = 1:7, directional = 1, row.col = 1:5) # 设置弦方向为从行名到列名，设置弦颜色
 > circos.clear()

 > library(circlize)
 > 
 > arr_col <- data.frame(c("S1", "S2", "S3"), c("E5", "E6", "E4"),
 +                       c("black", "black", "black")) # 生成箭头的颜色3列特征数据框
 > chordDiagram(mat, grid.col = grid_col, directional = 1,
 +              link.arr.col = arr_col, direction.type = "arrows", link.arr.length = 0.2)
 > circos.clear()
 > 
 > # 同时设置箭头和弦高diffHeight
 > chordDiagram(mat, grid.col = grid_col, directional = 1,
 +     direction.type = c("diffHeight", "arrows"), # 同时设置箭头和弦高
 +     link.arr.col = arr_col, link.arr.length = 0.2)
 > circos.clear()
 > 
 > par(mfrow = c(1, 2))
 > # 指定箭头类型为大箭头
 > matx <-  matrix(rnorm(64), 8)
 > 
 > chordDiagram(matx, directional = 1, direction.type = c("diffHeight", "arrows"),
 +     link.arr.type = "big.arrow") # 大箭头，箭头和箭杆合二为一
 > circos.clear()
 > 
 > # 大箭头加调整弦高diffHeight
 > chordDiagram(matx, directional = 1, direction.type = c("diffHeight", "arrows"),
 +     link.arr.type = "big.arrow", diffHeight = -uh(2, "mm")) # 设置弦高为-2mm
 > circos.clear()

 > library(circlize)
 > 
 > # 默认移除小比例值
 > mat <- matrix(rnorm(36), 6, 6)
 > rownames(mat) <-  paste0("R", 1:6)
 > colnames(mat) <- paste0("C", 1:6)
 > mat[2, ] <- 1e-10 # 将第2行所有值改成很小的值
 > mat[, 3] <- 1e-10 # 将第3列所有值改成很小的值
 > 
 > chordDiagram(mat)
 > circos.info() # 显示绘图的对象，不包含第2行的行名(R2)和第3列的列名(C3)，则表示被移除了
 All your sectors:
  [1] "R1" "R3" "R4" "R5" "R6" "C1" "C2" "C4" "C5" "C6"
 
 All your tracks:
 [1] 1 2
 
 Your current sector.index is C6
 Your current track.index is 2
 > circos.clear()
 > 
 > # reduce参数调整
 > mat[2, ] <- 1e-2
 > chordDiagram(mat, reduce = 1e-3) # 控制reduce参数比C2小，则C2行不会被移除
 > circos.info()
 All your sectors:
  [1] "R1" "R2" "R3" "R4" "R5" "R6" "C1" "C2" "C4" "C5" "C6"
 
 All your tracks:
 [1] 1 2
 
 Your current sector.index is C6
 Your current track.index is 2
 > circos.clear()

 > library(circlize)
 > 
 > par(mfrow = c(1, 3)) # 多图布局，分3列排版
 > chordDiagram(mat, grid.col = grid_col, annotationTrack = "grid") # 只显示网格，不显示刻度线和标签轨道
 > chordDiagram(mat, grid.col = grid_col, annotationTrack = c("name", "grid"), # 指定显示标签和网格轨道
 +     annotationTrackHeight = c(0.03, 0.01)) # 指定标签轨道和网格轨道的环高 
 > 
 > chordDiagram(mat, grid.col = grid_col, annotationTrack = NULL) # 移除所有轨道
 > circos.clear()

 ## Data
 # read the data
 mat <- read.csv("D:/Pathway-Disease.csv",
                 header = T,
                 row.names = 1,
                 stringsAsFactors = F)
 # converted to matrix
 mat <- as.matrix(mat)
 # drop small flows, here: quantile(mat,0.65)=3, any element <=3 is drop
 mat[mat<=quantile(mat,0.65)] <-0
 
 ## Chord Siagram
 # load circlize
 library("circlize")
 # setting reset
 circos.clear()
 
 # Open a window interface
 #win.graph(width=20, height=20,pointsize=12)
 
 # plot chord siagram
 chordDiagram(mat,
              annotationTrack = "grid",
              preAllocateTracks = 1)
 
 #chordDiagram(mat,
 #             directional = 1,
 #             annotationTrack = "grid",
 #             annotationTrackHeight = c(0.1, 0.1),
 #             diffHeight  = -0.04,
 #             preAllocateTracks = 1)
 
 # add labels
 circos.track(track.index = 1, panel.fun = function(x, y) {
   circos.text(CELL_META$xcenter, CELL_META$ylim[1], CELL_META$sector.index,
               facing = "clockwise", niceFacing = TRUE, adj = c(0, 0.5), cex = 1)
 }, bg.border = NA)
 
 # save the image as a pdf file