pangxie的个人博客分享 http://blog.sciencenet.cn/u/pangxie

博文

heatmap热图绘制教程

已有 9855 次阅读 2021-1-10 15:12 |个人分类:R语言|系统分类:科研笔记

heatmap.csv

#教程借鉴于R包pheatmap:用参数一步步详细绘制热图 – 我的工作墙 (gzxuexi.xyz)

#第一步:公众路径设置,调用包pheatmap,读取目的文件,查看文件。

rm(list=ls())

#setwd("")

#getwd()

library(pheatmap)

library(ggplot2)

data <- read.csv("heatmap.csv",header=T,row.names=1)

dim(data)

head(data)

#第二步:逐步深入绘制热图(图1-图15)

#一建热图(图1)

p1.1<-pheatmap(data)

#设置标准化方向scale,对其横向标准化

p1.2<-pheatmap(data,scale="row")

#设置边框为白色,横向纵向聚类为无;border="white;

cluster_cols = F;cluster_rows = F

p1.3<-pheatmap(data,scale="row",border="white",cluster_cols = F,cluster_rows = F)

#去掉横纵坐标中的id;show_rownames = F,show_colnames = F(图2)

p2<-pheatmap(data,scale="row",show_rownames = F,show_colnames = F,

            border="white",cluster_cols = F,cluster_rows = F)

#去掉右上角图例;legend = F(图3)

p3<-pheatmap(data,scale="row",show_rownames = F,show_colnames = F,legend = F,

            border="white",cluster_cols = F,cluster_rows = F)

#设置右上角图例的范围;legend_breaks=c(-1,1)(图4)

p4.1<-pheatmap(data,scale="row",legend_breaks=c(-1,0,1),

            show_rownames = F,show_colnames = F,legend = T,border="white",

            cluster_cols = F,cluster_rows = F)

p4.2<-pheatmap(data,scale="row",legend_breaks=c(-2,0,2),

            show_rownames = F,show_colnames = F,legend = T,border="white",

            cluster_cols = F,cluster_rows = F)

#设置图中字的大小;fondsize(图5)

#图表字体:命令:fondsize=2/8

p5.1<-pheatmap(data,scale="row",fontsize = 2,show_rownames = T,

            show_colnames = T,legend = F,border="white",

            cluster_cols = F,cluster_rows = F)

p5.2<-pheatmap(data,scale="row",fontsize = 8,show_rownames = T,

             show_colnames = T,legend = F,border="white",

             cluster_cols = F,cluster_rows = F)

#改变横向和纵向字体大小;fontsize_row = 8,fontsize_col=12 (图6)

p6.1<-pheatmap(data,scale="row",fontsize_row = 8,fontsize_col = 12,

            show_rownames = T,show_colnames = T,legend=T,border="white",

            cluster_cols = T,cluster_rows = T)

p6.2<-pheatmap(data,scale="row",fontsize_row = 12,fontsize_col = 8,

            show_rownames = T,show_colnames = T,legend=T,border="white",

            cluster_cols = T,cluster_rows = T)

#设置横向纵向的树高;treeheight_col = 20,treeheight_row = 15(图7)

p7.1<-pheatmap(data,scale="row",border="white",cluster_cols = T,

            treeheight_col = 20,cluster_rows = T,treeheight_row = 15)

p7.2<-pheatmap(data,scale="row",border="white",cluster_cols = T,

            treeheight_col = 20,cluster_rows = T,treeheight_row = 20)

#设置聚类的距离类型clustering_distance_rows(图8)

#clustering_distance_rows分为如下几类

#'correlation', 'euclidean', 'maximum', 'manhattan', 'canberra', 

#'binary', 'minkowski'

p8.1<-pheatmap(data,scale="row",clustering_distance_rows = "correlation",

            border="white",cluster_cols = T,treeheight_col = 20,

            cluster_rows = T,treeheight_row = 20)

p8.2<-pheatmap(data,scale="row",clustering_distance_rows = "minkowski",

            border="white",cluster_cols = T,treeheight_col = 20,

            cluster_rows = T,treeheight_row = 20)

#调整聚类的方法;clustering_method="single"(图9)

#'ward', 'ward.D', 'ward.D2', 'single', 'complete', 'average', 

#''mcquitty', 'median' or 'centroid'.

p9.1<-pheatmap(data,scale="row",clustering_distance_rows = "minkowski",

            clustering_method="single",border="white",cluster_cols = T,

            treeheight_col = 20,cluster_rows = T,treeheight_row = 20)

p9.2<-pheatmap(data,scale="row",clustering_distance_rows = "minkowski",

            clustering_method="complete",border="white",cluster_cols = T,

            treeheight_col = 20,cluster_rows = T,treeheight_row = 20)

#分组类别类型,斜的?45°?(图10)

#angle_col=315

#可选“270”, “0”, “45”, “90”, “315”

p10.1<-pheatmap(data,scale="row",angle_col = 0,

            clustering_distance_rows = "minkowski",clustering_method="complete",

            border="white",cluster_cols = T,treeheight_col = 20,cluster_rows = T,

            treeheight_row = 20)

p10.2<-pheatmap(data,scale="row",angle_col = 45,

            clustering_distance_rows = "minkowski",clustering_method="complete",

            border="white",cluster_cols = T,treeheight_col = 20,cluster_rows = T,

            treeheight_row = 20)

#给图形增加标题;main="Gene1"(图11)

p11<-pheatmap(data,scale="row",main="Gene1",angle_col = 45,

            clustering_distance_rows = "minkowski",clustering_method="complete",

            border="white",cluster_cols = T,treeheight_col = 20,cluster_rows = T,

            treeheight_row = 20)

#调整热图方块宽度和高度,cellwidth = 8,cellheight = 6(图12)

p12.1<-pheatmap(data,scale="row",cellwidth = 6,cellheight = 6,main="Gene1",

            angle_col = 45,clustering_distance_rows = "minkowski",

            clustering_method="complete",border="white",cluster_cols = T,

            treeheight_col = 20,cluster_rows = T,treeheight_row = 20)

p12.2<-pheatmap(data,scale="row",cellwidth = 14,cellheight = 10,main="Gene1",

                angle_col = 45,clustering_distance_rows = "minkowski",

                clustering_method="complete",border="white",cluster_cols = T,

                treeheight_col = 20,cluster_rows = T,treeheight_row = 20)

#根据热图聚类对其进行区块儿划分(图13)

#cutree_cols = 3,cutree_rows =4

p13.1<-pheatmap(data,scale="row",cutree_cols = 3,cutree_rows =4,main="Gene1",

            angle_col = 0,clustering_distance_rows = "minkowski",

            clustering_method="complete",border="white",cluster_cols = T,

            treeheight_col = 20,cluster_rows = T,treeheight_row = 20)

#在图13.1基础上增加边缘线(图13.2)

p13.2<-pheatmap(data,scale="row",border="#8B0A50",cutree_cols = 3,

                cutree_rows =4,main="Gene1",angle_col = 0,

                clustering_distance_rows = "minkowski",

                clustering_method="complete",cluster_cols = T,

                treeheight_col = 20,cluster_rows = T,treeheight_row = 20)

#调整Gene的数值

#是否展示数值,大小和颜色,大小以及数值展示类型

#display_numbers = T,fontsize_number = 2,number_color="black"

#number_format:

p14.1<-pheatmap(data,scale="row",border="#8B0A50",display_numbers = T,

            cutree_cols = 3,cutree_rows =4,main="Gene1",angle_col = 0,

            clustering_distance_rows = "minkowski",clustering_method="complete",

            cluster_cols = T,treeheight_col = 20,cluster_rows = T,

            treeheight_row = 20)

p14.2<-pheatmap(data,scale="row",border="#8B0A50",fontsize_number = 2,

            display_numbers = T,cutree_cols = 3,cutree_rows =4,main="Gene1",

            angle_col = 0,clustering_distance_rows = "minkowski",

            clustering_method="complete",cluster_cols = T,treeheight_col = 20,

            cluster_rows = T,treeheight_row = 20)

p14.3<-pheatmap(data,scale="row",border="#8B0A50",fontsize_number = 8,

            display_numbers = T,cutree_cols = 3,cutree_rows =4,main="Gene1",

            angle_col = 0,clustering_distance_rows = "minkowski",

            clustering_method="complete",cluster_cols = T,treeheight_col = 20,

            cluster_rows = T,treeheight_row = 20)

p14.3<-pheatmap(data,scale="row",number_color="red",border="#8B0A50",

                fontsize_number = 8,display_numbers = T,cutree_cols = 3,

                cutree_rows =4,main="Gene1",angle_col = 0,

                clustering_distance_rows = "minkowski",

                clustering_method="complete",cluster_cols = T,

                treeheight_col = 20,cluster_rows = T,treeheight_row = 20)

p14.4<-pheatmap(data,scale="row",number_color="red",number_format="%.2e",

                border="#8B0A50",fontsize_number = 8,display_numbers = T,

                cutree_cols = 3,cutree_rows =4,main="Gene1",angle_col = 0,

                clustering_distance_rows = "minkowski",

                clustering_method="complete",cluster_cols = T,

                treeheight_col = 20,cluster_rows = T,treeheight_row = 20)

#对热图方块儿进行标记;display_numbers,如果该值大于10,则为+/**,否则为-/*(图15)

p15.1<-pheatmap(data,scale="row",number_color="red",number_format="%.2e",

            border="#8B0A50",fontsize_number = 8,display_numbers = 

              matrix(ifelse(data > 10, "+", "-"), nrow(data)),cutree_cols = 3,

            cutree_rows =4,main="Gene1",angle_col = 0,clustering_distance_rows =

              "minkowski",clustering_method="complete",cluster_cols = T,

            treeheight_col = 20,cluster_rows = T,treeheight_row = 20)

p15.2<-pheatmap(data,scale="row",number_color="red",number_format="%.2e",

            border="#8B0A50",fontsize_number = 8,display_numbers = 

              matrix(ifelse(data > 10, "**", "*"), nrow(data)),cutree_cols = 3,

            cutree_rows =4,main="Gene1",angle_col = 0,clustering_distance_rows 

            = "minkowski",clustering_method="complete",cluster_cols = T,

            treeheight_col = 20,cluster_rows = T,treeheight_row = 20)

#第三步:构建分组信息(图16-图18)

#构建纵向和横向分组信息(有3个处理,分别是:盐、干旱和热应激);

#以及时间:0-3day,对3类基因21个基因进行分组,分别是:”WRKY”, “AP2”, “YABBY”

#构建纵向分组信息(图16)

annotation_col = data.frame(Deal_with = 

                factor(rep(c("Salt", "Drought","Heat"), 3)),

                Day=factor(rep(c("Day1", "Day2","Day3"), 3)))

rownames(annotation_col)

colnames(data)

rownames(annotation_col) <- colnames(data)

rownames(annotation_col)

head(annotation_col)

#纵向分组热图

pheatmap(data, annotation_col = annotation_col)

p16<-pheatmap(data,scale="row",annotation_col = annotation_col,

              number_color="red",number_format="%.2e",border="#8B0A50",

              fontsize_number = 8,display_numbers =

              matrix(ifelse(data > 10, "**", "*"), nrow(data)),

              cutree_cols = 3,cutree_rows =4,main="Gene1",angle_col = 0,

              clustering_distance_rows = "minkowski",

              clustering_method="complete",cluster_cols = T,treeheight_col = 20,

              cluster_rows = T,treeheight_row = 20)

#横向分组信息(图17)

annotation_row = data.frame(GeneClass = factor(rep(c("WRKY", "AP2", "YABBY"),7)))

rownames(annotation_row) <- rownames(data)

head(annotation_row)

#横向分组热图

pheatmap(data, annotation_row =annotation_row)

p17<-pheatmap(data,scale="row",number_color="red", 

            annotation_row =annotation_row,number_format="%.2e",

            border="#8B0A50",fontsize_number = 8,

            display_numbers = matrix(ifelse(data > 10, "++", "-"), nrow(data)),

            cutree_cols = 3,cutree_rows =4,main="Gene1",angle_col = 0,

            clustering_distance_rows = "minkowski",clustering_method="complete",

            cluster_cols = T,treeheight_col = 20,cluster_rows = T,

            treeheight_row = 20)

#共同组合二者(图18)

pheatmap(data,annotation_col = annotation_col,

            annotation_row = annotation_row)

p18<-pheatmap(data,scale="row",number_color="red",

            annotation_col = annotation_col,annotation_row = annotation_row,

            number_format="%.2e",border="#8B0A50",fontsize_number = 8,

            display_numbers = matrix(ifelse(data > 10, "++", "-"), nrow(data)),

            cutree_cols = 3,cutree_rows =4,main="Gene1",angle_col = 0,

            clustering_distance_rows = "minkowski",clustering_method="complete",

            cluster_cols = T,treeheight_col = 20,cluster_rows = T,

            treeheight_row = 20)

setwd("C:/Users/53446/Desktop/heatmap")

ggsave(p1.1,filename = "p1.1.jpeg",width = 6,height = 8)

ggsave(p1.2,filename = "p1.2.jpeg",width = 6,height = 8)

ggsave(p1.3,filename = "p1.3.jpeg",width = 6,height = 8)

ggsave(p10.1,filename = "p10.1.jpeg",width = 6,height = 8)

ggsave(p10.2,filename = "p10.2.jpeg",width = 6,height = 8)

ggsave(p11,filename = "p11.jpeg",width = 6,height = 8)

ggsave(p12.1,filename = "p12.1.jpeg",width = 6,height = 8)

ggsave(p12.2,filename = "p12.2.jpeg",width = 6,height = 8)

ggsave(p13.1,filename = "p13.1.jpeg",width = 6,height = 8)

ggsave(p13.2,filename = "p13.2.jpeg",width = 6,height = 8)

ggsave(p14.1,filename = "p14.1.jpeg",width = 6,height = 8)

ggsave(p14.2,filename = "p14.2.jpeg",width = 6,height = 8)

ggsave(p14.3,filename = "p14.3.jpeg",width = 6,height = 8)

ggsave(p14.4,filename = "p14.4.jpeg",width = 6,height = 8)

ggsave(p15.1,filename = "p15.1.jpeg",width = 6,height = 8)

ggsave(p15.2,filename = "p15.2.jpeg",width = 6,height = 8)

ggsave(p16,filename = "p16.jpeg",width = 6,height = 8)

ggsave(p17,filename = "p17.jpeg",width = 6,height = 8)

ggsave(p18,filename = "p18.jpeg",width = 6,height = 8)

ggsave(p2,filename = "p2.jpeg",width = 6,height = 8)

ggsave(p3,filename = "p3.jpeg",width = 6,height = 8)

ggsave(p4.1,filename = "p4.1.jpeg",width = 6,height = 8)

ggsave(p4.2,filename = "p4.2.jpeg",width = 6,height = 8)

ggsave(p5.1,filename = "p5.1.jpeg",width = 6,height = 8)

ggsave(p5.2,filename = "p5.2.jpeg",width = 6,height = 8)

ggsave(p6.1,filename = "p6.1.jpeg",width = 6,height = 8)

ggsave(p6.2,filename = "p6.2.jpeg",width = 6,height = 8)

ggsave(p7.1,filename = "p7.1.jpeg",width = 6,height = 8)

ggsave(p7.2,filename = "p7.2.jpeg",width = 6,height = 8)

ggsave(p8.1,filename = "p8.1.jpeg",width = 6,height = 8)

ggsave(p8.2,filename = "p8.2.jpeg",width = 6,height = 8)

ggsave(p9.1,filename = "p9.1.jpeg",width = 6,height = 8)

ggsave(p9.2,filename = "p9.2.jpeg",width = 6,height = 8)











pheatmap(mat, color = colorRampPalette(rev(brewer.pal(n = 7, name =

                                                        "RdYlBu")))(100), kmeans_k = NA, breaks = NA, border_color = "grey60",

         cellwidth = NA, cellheight = NA, scale = "none", cluster_rows = TRUE,

         cluster_cols = TRUE, clustering_distance_rows = "euclidean",

         clustering_distance_cols = "euclidean", clustering_method = "complete",

         clustering_callback = identity2, cutree_rows = NA, cutree_cols = NA,

         treeheight_row = ifelse((class(cluster_rows) == "hclust") || cluster_rows,

                                 50, 0), treeheight_col = ifelse((class(cluster_cols) == "hclust") ||

                                                                   cluster_cols, 50, 0), legend = TRUE, legend_breaks = NA,

         legend_labels = NA, annotation_row = NA, annotation_col = NA,

         annotation = NA, annotation_colors = NA, annotation_legend = TRUE,

         annotation_names_row = TRUE, annotation_names_col = TRUE,

         drop_levels = TRUE, show_rownames = T, show_colnames = T, main = NA,

         fontsize = 10, fontsize_row = fontsize, fontsize_col = fontsize,

         angle_col = c("270", "0", "45", "90", "315"), display_numbers = F,

         number_format = "%.2f", number_color = "grey30", fontsize_number = 0.8

         * fontsize, gaps_row = NULL, gaps_col = NULL, labels_row = NULL,

         labels_col = NULL, filename = NA, width = NA, height = NA,

         silent = FALSE, na_col = "#DDDDDD", ...)


## 函数解释

mat # 需要绘制热图的数字矩阵。

color # 表示颜色,赋值渐变颜色调色板colorRampPalette属性,选择"蓝,白,红"渐变,分为100个等级

# 例 color =  colorRampPalette(c("navy", "white", "red"))(100)

kmeans_k # 绘制热图的行聚类数,如果是NA,那么行不会聚类。

breaks  # 设置mat数值范围的数字序列

border_color # 表示热图上单元格边框的颜色,如果不绘制边框,则使用NA

cellwidth # 表示每个单元格的宽度,若选择NA则表示适应窗口

cellheight # 表示每个单元格的高度,若选择NA则表示适应窗口

scale # 表示值均一化的方向,或者按照行或列,或者没有,值可以是"row",  “column” 或者"none"

cluster_rows # 表示进行行的聚类,值可以是FALSE或TRUE

cluster_cols # 表示进行列的聚类,值可以是FALSE或TRUE

clustering_distance_rows # 表示行距离度量的方法

clustering_distance_cols # 表示列距离度量的方法

clustering_method # 表示聚类方法,值可以是hclust的任何一种,

# 如"ward.D",“single”,  “complete”, “average”, “mcquitty”, “median”, “centroid”, “ward.D2”

clustering_callback # 回调函数来修改聚类

cutree_rows # 基于层次聚类(使用cutree)划分行的簇数(如果未聚集行,则忽略参数)

cutree_cols  # 基于层次聚类(使用cutree)划分列的簇数(如果未聚集行,则忽略参数)

treeheight_row # 行的树的高度,

treeheight_col # 列的树的高度

legend # TRUE或者FALSE,表示是否显示图例

legend_breaks # 设置图例的断点,如legend_breaks =  -1:4

legend_labels # legend_breaks对应的标签例:legend_breaks = -1:4, legend_labels = c(“0”,“1e-4”, “1e-3”, “1e-2”, “1e-1”, “1”)

annotation_row # 行的分组信息,需要使用相应的行名称来匹配数据和注释中的行,注意之后颜色设置会考虑离散值还是连续值,格式要求为数据框

annotation_col # 列的分组信息,需要使用相应的列名称来匹配数据和注释中的列,格式要求为数据框

annotation_colors  # 用于手动指定annotation_row和annotation_col  track颜色的列表。

annotation_legend # 是否显示图例的名称。

annotation_names_row # 是否显示行注释的名称。

annotation_names_col # 是否显示列注释的名称。

show_rownames # 是否显示行名

show_colnames # 是否显示列名

main  # 图的名字

fontsize # 图的字体大小

fontsize_row # 行名的字体大小,默认与图的字体大小相同

fontsize_col # 列名的字体大小,默认与图的字体大小相同

angle_col # 列标签的角度,可选择 (0, 45, 90, 270 and 315)

display_numbers # 表示是否将数值显示在热图的格子中,如果这是一个矩阵(与原始矩阵具有相同的尺寸),则显示矩阵的内容而不是原始值。

number_format # 设置显示数值的格式,较常用的有"%.2f"(保留小数点后两位),"%.1e"(科学计数法显示,保留小数点后一位)

number_color # 设置显示内容的颜色

fontsize_number # 设置显示内容的字体大小

labels_row  # 代替行名的自定义标签 

labels_col # 代替列名的自定义标签

filename # 图片保存位置以及文件名

width # 手动设置输出文件的宽度(单位:英寸)

height # 手动设置输出文件的高度(单位:英寸)

silent # 不绘制热图

na_col # 缺失值的颜色


#计算相关性

cor = corr.test(lefse, plant, method="spearman", adjust="none")

#提取相关性、p

cmt <-cor$r

pmt <- cor$p

#判断显著性

if (!is.null(pmt)){

ssmt <- pmt< 0.01

pmt[ssmt] <-'**'

smt <- pmt >0.01& pmt <0.05

pmt[smt] <- '*'

pmt[!ssmt&!smt]<- ''

} else {

  pmt <- F

 }

#自定义颜色范围

mycol<-colorRampPalette(c("blue","white","tomato"))(800)

 #可视化

pheatmap(cmt,scale = "none", cluster_row = F, cluster_col = F, border=NA, fontsize_row=8, fontsize_col=8, display_numbers = pmt, fontsize_number = 10, number_color = "white", cellwidth = 13, cellheight =13,color=mycol, filename="plant.lefse.jpg")





https://blog.sciencenet.cn/blog-3448646-1266557.html

上一篇:差异分析完整解决方案:Easystat
下一篇:双歧杆菌blast本地教程
收藏 IP: 122.192.12.*| 热度|

0

该博文允许注册用户评论 请点击登录 评论 (0 个评论)

数据加载中...
扫一扫,分享此博文

全部作者的其他最新博文

Archiver|手机版|科学网 ( 京ICP备07017567号-12 )

GMT+8, 2024-12-24 11:48

Powered by ScienceNet.cn

Copyright © 2007- 中国科学报社

返回顶部