php热图,科学网—使用ComplexHeatmap包绘制个性化热图 - 刘永鑫的博文

最新推荐文章于 2024-07-24 15:29:57 发布

周远凡

最新推荐文章于 2024-07-24 15:29:57 发布

阅读量469

点赞数

文章标签： php热图

使用ComplexHeatmap包绘制个性化热图

作者：刘梦瑶诺禾致源微生物信息

审稿：刘永鑫中国科学院遗传与发育生物学研究所

ComplexHeatmap包由顾祖光博士创建，是一个非常全面的绘制热图的R包，可以利用它来绘制许多文献中的美图，例如下图展示的16S文献分析中的热图。这里主要介绍一下如何用这个R包来绘制类似的个性化热图。

检测安装加载包# 检测安装CRAN包

package_list = c("circlize","grid","BiocManager")

for(p in package_list){

if (!requireNamespace(p, quietly = TRUE))

install.packages(p)

}

# 检测安装bioconductor包

package_list = c("ComplexHeatmap")

for(p in package_list){

if (!requireNamespace(p, quietly = TRUE))

BiocManager::install(p)

}

# 加载依赖包

library(circlize)

library(grid)

library(ComplexHeatmap)

创建测试数据集# 设置随机数种子，保证数据分析随机过程可重复

set.seed(123)

# 生成模拟数据：12行成10列矩阵

mat = cbind(rbind(matrix(rnorm(16, -1), 4),

matrix(rnorm(32, 1), 8)),

rbind(matrix(rnorm(24, 1), 4),

matrix(rnorm(48, -1), 8)))

# 随机重排

mat = mat[sample(nrow(mat), nrow(mat)),

sample(ncol(mat), ncol(mat))]

# 添加行、列名

rownames(mat) = paste0("R", 1:12)

colnames(mat) = paste0("C", 1:10)

一行命令绘图

使用默认参数，一行命令即可出图#默认对行和列都进行聚类

Heatmap(mat)

调参美化

下面我们通过参数设置来进行个性化热图定制。

使用HeatmapAnnotation函数可以构建注释对象，我们可以进行自定义，也可以直接使用它的内置函数。

注释按位置来分类可分为行注释和列注释，以列注释为例，其内置函数按照图形的类型可以分为6种，anno_points()，anno_barplot()，anno_boxplot()，anno_histogram()，anno_density()，anno_text()。

本文重点讨论anno_points()的用法。# 生成包含10个0.5数值的向量

value = rep(0.5,10)

# 设置值、形状、大小、颜色等

ha = HeatmapAnnotation("type" = anno_points(value, pch=c(19,19,15,15,24,24,23,23,3,3), size = unit(7, "mm"),gp = gpar(col = c("#bf94e4","#bf94e4","#bf94e4","#bf94e4","#1dacd6","#1dacd6","#1dacd6","#1dacd6","red","red")),border=FALSE,ylim=c(0,1)),show_annotation_name = FALSE)

“type”为这一行注释的名称，show_annotation_name = FALSE，即不显示名称。pch可指定绘制点时使用的符号，共25种，如上三角，下三角，圆形，方形等，具体可见《R In Action》。size可指定符号的大小，gp可指定符号的颜色。# 批量按行中心标准化，减均值除方差，Z-score

mat_scaled = apply(mat, 1, scale)

# 继续原数据表列名

rownames(mat_scaled) = colnames(mat)

# 转置才与原方向一致

mat_scaled = t(mat_scaled)

# 通过circlize包中的colorRamp2()函数，来自定义颜色

col_fun = circlize::colorRamp2(c(-3, 0, 3), c("black", "white", "yellow"))

# 新矩阵

shape

# 循环元素筛选，变为+或空，显著标记常用

for(i in 1:x ){

for (j in 1:y ){

if(shape[i,j]>=1){

shape[i,j]

}}}

如需对数据进行标准化，需使用apply函数来处理数据。我们可以通过circlize包中的colorRamp2()函数，来自定义颜色。对mat_scaled的数值进行筛选，生成一个符号是加号或空值的新数据框。这一部分可以根据作图要求来自定义。P1=Heatmap(mat_scaled,

name = "hello",

top_annotation = ha,

col = col_fun,

rect_gp = gpar(col = "black",lty = 2, lwd = 1),

cell_fun = function(j, i, x, y, width, height, fill) {grid.text(shape[i,j], x = x, y = y,gp = gpar(fontsize = 10,col="red"))},

cluster_rows = FALSE,

cluster_columns = FALSE,

row_names_side = "left",

column_names_side="bottom",

row_names_gp = gpar(col = c("#8B7500","#8B7500","#8B7500","#8B7500","#8B7500","#8B7500","#0000FF","#0000FF","#0000FF","#0000FF","#0000FF","#0000FF")))

name可定义图例的名称。top_annotation 可引用上面定义好的列注释, 并将列注释放在heatmap上方；bottom_annotation 则将列注释放在heatmap下方。rect_gp定义小方格的边框颜色，线条类型及宽度。cell_fun可以对heatmap的每个小方格进行自定义，这里用其来显示”+”号，也可以显示数字等。cluster_rows和cluster_columns可定义是否聚类。row_names_side可定义行名的显示位置，默认值right。column_names_side可定义列名的显示位置，默认值bottom。row_names_gp可定义列名的颜色。# 行名第一列

texta = c("A","B","C","D","EEEEE","F","G","H","I","J","K","L")

# 行注释，宽度为最大文本

ha_texta =rowAnnotation(text = row_anno_text(texta), width = max_text_width(texta))

# 行名第二列

textb = c("M","N","O","P","Q","R","S","T","U","V","W","X")

ha_textb =rowAnnotation(text = row_anno_text(textb), width = max_text_width(textb))

# 添加行名注释

ht_list = P1 + ha_texta + ha_textb

# 添加图例对应文字、形状和颜色

lgd = legendGrob(c("A","B","C","D","E"), pch = c(19,15,24,23,3),gp= gpar(col =c("#bf94e4","#bf94e4","#1dacd6","#1dacd6","red")))

# 绘图，添加热图图例左，注释图例

draw(ht_list,heatmap_legend_side = "left",annotation_legend_list = list(lgd))