提供了编程的基础技术教程

zl程序教程

您现在的位置是：首页 >  数据库

当前栏目

R语言ggplot2科研数据数据可视化实用手册~第八章热图（heatmap）

2023-02-19 12:27:49 时间

Chapter 8 R语言ggplot2热图

今天下午7点到9点直播讲解如下代码，腾讯会议，感兴趣的参加，给推文打赏10元获取腾讯会议直播链接

8.1 pheatmap热图

R语言里做热图最快捷的方式是用pheatmap这个R包,优点是用非常少的代码就可以出一个比较好看的图，缺点是细节修改不是很方便，比如要用热图展示基因表达量的数据，准备数据的格式如下

pheatmap不是R语言自带的R包，第一次使用需要先安装,安装直接使用命令install.packages("pheatmap")

读取数据

library(readxl)dat01<-read_excel("example_data/06-lineplot/dat08.xlsx")

这里需要注意的 一个点是热图数据通常需要把第一列的基因名作为整个数据的行名，但是读取excel的函数好像没有指定列为行名的函数，当然可以将数据集读取进来以后再进行转换，另外一种方式就是把数据另存为csv格式，然后用读取csv格式数据的函数

这里需要注意的一点是转换为csv格式的时候选择如下截图中的2，不要选择1

读取csv格式数据

dat01<-read.delim(file = "example_data/08-heatmap/01pheatmap_example.csv",                  header=TRUE,                  sep=",",                  row.names = 1)

作图代码

dat01<-read.delim(file = "example_data/08-heatmap/01pheatmap_example.csv",                  header=TRUE,                  sep=",",                  row.names = 1)library(pheatmap)pheatmap(dat01)

如果要在色块上添加文本，再单独准备一个和热图数据格式一样的数据，然后用display_numbers参数添加文本,这里我就直接使用热图的数据

dat01<-read.delim(file = "example_data/08-heatmap/01pheatmap_example.csv",                  header=TRUE,                  sep=",",                  row.names = 1)library(pheatmap)pheatmap(dat01)
pheatmap(dat01,         display_numbers = dat01)

以上是对pheatmap这个R包的简要介绍

ggplot2也有直接做热图的函数 geom_tile(),ggplot2做热图可能代码稍微繁琐，但是优点是细节调整方便，基本上所有的细节都可以用代码来调整

ggplot2做热图还需要掌握的一个知识点是 长格式数据 和 宽格式 数据，ggplot2作图的输入数据都是长格式数据，长格式数据如下，一列x，一列y，还有一个数据

宽格式数据截图如下

这个长宽格式转化是ggplot2作图必须理解的一个概念

R语言里提供了长宽格式数据互相转化的函数，这里我以tidyverse这个R包里的函数作为介绍，tidyverse主要是用来在数据处理的，也不是R语言自带的R包，需要运行安装命令install.packages("tidyverse")

宽格式数据转换为长格式用到的函数是pivot_longer()

转换代码

library(readxl)dat01<-read_excel("example_data/08-heatmap/02_wide_data.xlsx")head(dat01)
## # A tibble: 2 x 3
##   gene_name sample1 sample2
##   <chr>       <dbl>   <dbl>
## 1 gene1           4       3
## 2 gene2           5       4
library(tidyverse)dat01 %>%   pivot_longer(-gene_name,names_to = "A",values_to = "B")
## # A tibble: 4 x 3
##   gene_name A           B
##   <chr>     <chr>   <dbl>
## 1 gene1     sample1     4
## 2 gene1     sample2     3
## 3 gene2     sample1     5
## 4 gene2     sample2     4

长格式转换为宽格式的函数是pivot_wider()

转换代码

library(readxl)dat01<-read_excel("example_data/08-heatmap/02_long_data.xlsx")head(dat01)
## # A tibble: 4 x 3
##   x     y       value
##   <chr> <chr>   <dbl>
## 1 gene1 sample1     3
## 2 gene1 sample2     4
## 3 gene2 sample1     5
## 4 gene2 sample2     6
library(tidyverse)dat01 %>%   pivot_wider(names_from = y,values_from = value)
## # A tibble: 2 x 3
##   x     sample1 sample2
##   <chr>   <dbl>   <dbl>
## 1 gene1       3       4
## 2 gene2       5       6

这个是最基本的长宽格式数据转换，如果数据集有很多列，有时候转换会相对比较复杂，这里就不做介绍，因为我也搞不懂有时候

8.2 ggplot2热图

以下介绍ggplot2做热图的代码都是假设已经拿到了长格式数据

示例数据如下

最基本的热图代码

library(readxl)dat01<-read_excel("example_data/08-heatmap/03_heatmap_example.xlsx")head(dat01)
## # A tibble: 6 x 3
##   gene_name Sample   expr_value
##   <chr>     <chr>         <dbl>
## 1 gene_1    sample_1          1
## 2 gene_1    sample_2          5
## 3 gene_1    sample_3          8
## 4 gene_1    sample_4          4
## 5 gene_1    sample_5          2
## 6 gene_1    sample_6          5
library(ggplot2)ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color="red")
ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color=NA)

热图经常遇到的操作是调整坐标轴的顺序，这个可以通过赋予因子水平来实现

library(readxl)dat01<-read_excel("example_data/08-heatmap/03_heatmap_example.xlsx")head(dat01)
## # A tibble: 6 x 3
##   gene_name Sample   expr_value
##   <chr>     <chr>         <dbl>
## 1 gene_1    sample_1          1
## 2 gene_1    sample_2          5
## 3 gene_1    sample_3          8
## 4 gene_1    sample_4          4
## 5 gene_1    sample_5          2
## 6 gene_1    sample_6          5
dat01$gene_name<-factor(dat01$gene_name,                        levels = c("gene_1","gene_2","gene_3",                                   "gene_4","gene_5","gene_6",                                   "gene_7","gene_8","gene_9",                                   "gene_10"))library(ggplot2)ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color="red")
dat01$Sample<-factor(dat01$Sample,                     levels = c("sample_2","sample_4","sample_6",                                "sample_1","sample_3","sample_5"))ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color="red")

更改热图色块填充的颜色

更改热图填充颜色有很多种方式，这里我介绍我自己最常用的一种方式

参考链接

https://r-charts.com/

这里用到额外的一个R包 paletteer

https://github.com/EmilHvitfeldt/paletteer

library(readxl)dat01<-read_excel("example_data/08-heatmap/03_heatmap_example.xlsx")head(dat01)
## # A tibble: 6 x 3
##   gene_name Sample   expr_value
##   <chr>     <chr>         <dbl>
## 1 gene_1    sample_1          1
## 2 gene_1    sample_2          5
## 3 gene_1    sample_3          8
## 4 gene_1    sample_4          4
## 5 gene_1    sample_5          2
## 6 gene_1    sample_6          5
dat01$gene_name<-factor(dat01$gene_name,                        levels = c("gene_1","gene_2","gene_3",                                   "gene_4","gene_5","gene_6",                                   "gene_7","gene_8","gene_9",                                   "gene_10"))library(ggplot2)ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color="red")
library(paletteer)ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color="red")+  scale_fill_paletteer_c("ggthemes::Classic Orange-White-Blue")
ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color="red")+  scale_fill_paletteer_c("ggthemes::Classic Orange-White-Blue",                         direction = -1)

调整坐标轴文本标签的位置，y轴左右，x轴是上下

library(readxl)dat01<-read_excel("example_data/08-heatmap/03_heatmap_example.xlsx")head(dat01)
## # A tibble: 6 x 3
##   gene_name Sample   expr_value
##   <chr>     <chr>         <dbl>
## 1 gene_1    sample_1          1
## 2 gene_1    sample_2          5
## 3 gene_1    sample_3          8
## 4 gene_1    sample_4          4
## 5 gene_1    sample_5          2
## 6 gene_1    sample_6          5
dat01$gene_name<-factor(dat01$gene_name,                        levels = c("gene_1","gene_2","gene_3",                                   "gene_4","gene_5","gene_6",                                   "gene_7","gene_8","gene_9",                                   "gene_10"))library(ggplot2)library(paletteer)ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color="red")+  scale_fill_paletteer_c("ggthemes::Classic Orange-White-Blue",                         direction = -1)+  scale_x_discrete(position = "top")+  scale_y_discrete(position = "right")

调整坐标轴的文本方向

library(readxl)dat01<-read_excel("example_data/08-heatmap/03_heatmap_example.xlsx")head(dat01)
## # A tibble: 6 x 3
##   gene_name Sample   expr_value
##   <chr>     <chr>         <dbl>
## 1 gene_1    sample_1          1
## 2 gene_1    sample_2          5
## 3 gene_1    sample_3          8
## 4 gene_1    sample_4          4
## 5 gene_1    sample_5          2
## 6 gene_1    sample_6          5
dat01$gene_name<-factor(dat01$gene_name,                        levels = c("gene_1","gene_2","gene_3",                                   "gene_4","gene_5","gene_6",                                   "gene_7","gene_8","gene_9",                                   "gene_10"))library(ggplot2)library(paletteer)ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color="red")+  scale_fill_paletteer_c("ggthemes::Classic Orange-White-Blue",                         direction = -1)+  scale_x_discrete(position = "top")+  scale_y_discrete(position = "right")+  theme(axis.text.x = element_text(angle=60))
ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color="red")+  scale_fill_paletteer_c("ggthemes::Classic Orange-White-Blue",                         direction = -1)+  scale_x_discrete(position = "top")+  scale_y_discrete(position = "right")+  theme(axis.text.x = element_text(angle=60,hjust=0,vjust=1))

去掉整个的灰色背景和坐标轴的小短线

library(readxl)dat01<-read_excel("example_data/08-heatmap/03_heatmap_example.xlsx")head(dat01)
## # A tibble: 6 x 3
##   gene_name Sample   expr_value
##   <chr>     <chr>         <dbl>
## 1 gene_1    sample_1          1
## 2 gene_1    sample_2          5
## 3 gene_1    sample_3          8
## 4 gene_1    sample_4          4
## 5 gene_1    sample_5          2
## 6 gene_1    sample_6          5
dat01$gene_name<-factor(dat01$gene_name,                        levels = c("gene_1","gene_2","gene_3",                                   "gene_4","gene_5","gene_6",                                   "gene_7","gene_8","gene_9",                                   "gene_10"))library(ggplot2)library(paletteer)ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color="red")+  scale_fill_paletteer_c("ggthemes::Classic Orange-White-Blue",                         direction = -1)+  scale_x_discrete(position = "top")+  scale_y_discrete(position = "right")+  theme(axis.text.x = element_text(angle=60,hjust=0,vjust=1),        axis.ticks = element_blank(),        panel.background = element_blank())

如果看起来坐标轴的文本距离图比较远的话还可以调整

library(readxl)dat01<-read_excel("example_data/08-heatmap/03_heatmap_example.xlsx")head(dat01)
## # A tibble: 6 x 3
##   gene_name Sample   expr_value
##   <chr>     <chr>         <dbl>
## 1 gene_1    sample_1          1
## 2 gene_1    sample_2          5
## 3 gene_1    sample_3          8
## 4 gene_1    sample_4          4
## 5 gene_1    sample_5          2
## 6 gene_1    sample_6          5
dat01$gene_name<-factor(dat01$gene_name,                        levels = c("gene_1","gene_2","gene_3",                                   "gene_4","gene_5","gene_6",                                   "gene_7","gene_8","gene_9",                                   "gene_10"))library(ggplot2)library(paletteer)ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color="red")+  scale_fill_paletteer_c("ggthemes::Classic Orange-White-Blue",                         direction = -1)+  scale_x_discrete(position = "top",                   expand = expansion(mult = c(0,0)))+  scale_y_discrete(position = "right",                   expand = expansion(mult = c(0,0)))+  theme(axis.text.x = element_text(angle=60,hjust=0,vjust=1),        axis.ticks = element_blank(),        panel.background = element_blank())

在色块上添加文本，用geom_text()函数来添加

library(readxl)dat01<-read_excel("example_data/08-heatmap/03_heatmap_example.xlsx")head(dat01)
## # A tibble: 6 x 3
##   gene_name Sample   expr_value
##   <chr>     <chr>         <dbl>
## 1 gene_1    sample_1          1
## 2 gene_1    sample_2          5
## 3 gene_1    sample_3          8
## 4 gene_1    sample_4          4
## 5 gene_1    sample_5          2
## 6 gene_1    sample_6          5
dat01$gene_name<-factor(dat01$gene_name,                        levels = c("gene_1","gene_2","gene_3",                                   "gene_4","gene_5","gene_6",                                   "gene_7","gene_8","gene_9",                                   "gene_10"))library(ggplot2)library(paletteer)ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color="red")+  scale_fill_paletteer_c("ggthemes::Classic Orange-White-Blue",                         direction = -1)+  scale_x_discrete(position = "top",                   expand = expansion(mult = c(0,0)))+  scale_y_discrete(position = "right",                   expand = expansion(mult = c(0,0)))+  theme(axis.text.x = element_text(angle=60,hjust=0,vjust=1),        axis.ticks = element_blank(),        panel.background = element_blank())+  geom_text(aes(label=expr_value))

调整图例的细节

参考公众号推文 ggplot2画热图展示相关系数的简单小例子

截断和标签是在scale_fill函数里设置breaks和labels

图例的位置是在主题里进行设置

其他一些细节在guides函数里设置

library(readxl)dat01<-read_excel("example_data/08-heatmap/03_heatmap_example.xlsx")head(dat01)
## # A tibble: 6 x 3
##   gene_name Sample   expr_value
##   <chr>     <chr>         <dbl>
## 1 gene_1    sample_1          1
## 2 gene_1    sample_2          5
## 3 gene_1    sample_3          8
## 4 gene_1    sample_4          4
## 5 gene_1    sample_5          2
## 6 gene_1    sample_6          5
dat01$gene_name<-factor(dat01$gene_name,                        levels = c("gene_1","gene_2","gene_3",                                   "gene_4","gene_5","gene_6",                                   "gene_7","gene_8","gene_9",                                   "gene_10"))library(ggplot2)library(paletteer)ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr_value),color="red")+  scale_fill_paletteer_c("ggthemes::Classic Orange-White-Blue",                         direction = -1,                         breaks=c(1.1,3,5,7,9,9.9),                         labels=c(1,3,5,7,9,10))+  scale_x_discrete(position = "top",                   expand = expansion(mult = c(0,0)))+  scale_y_discrete(position = "right",                   expand = expansion(mult = c(0,0)))+  theme(axis.text.x = element_text(angle=60,hjust=0,vjust=1),        axis.ticks = element_blank(),        panel.background = element_blank())+  geom_text(aes(label=expr_value))+  theme(legend.position = "top")+  guides(fill=guide_colorbar(title = "AAA",                             title.position = "top",                             title.hjust = 0.5,                             barwidth = 20,                             ticks = FALSE,                             label = TRUE))

以上介绍的用来填充颜色的数据是连续型的，数据是离散的也是可以的，比如只关心某个基因在样本中是否表达，并不关心这个基因的表达量高低,示例数据如下

这里A代表基因表达B代表基因不表达，这个AB可以用任意字符代替

library(readxl)dat01<-read_excel("example_data/08-heatmap/04_heatmap_example.xlsx")head(dat01)
## # A tibble: 6 x 3
##   gene_name Sample   expr 
##   <chr>     <chr>    <chr>
## 1 gene_1    sample_1 B    
## 2 gene_1    sample_2 B    
## 3 gene_1    sample_3 B    
## 4 gene_1    sample_4 A    
## 5 gene_1    sample_5 A    
## 6 gene_1    sample_6 A
dat01$gene_name<-factor(dat01$gene_name,                        levels = c("gene_1","gene_2","gene_3",                                   "gene_4","gene_5","gene_6",                                   "gene_7","gene_8","gene_9",                                   "gene_10"))library(ggplot2)ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr),color="blue")
ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr),color="blue")+  scale_fill_manual(values = c("A"="#fe0904",                               "B"='#f9b54f'))
ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_tile(aes(fill=expr),color="blue")+  scale_fill_manual(values = c("A"="#fe0904",                               "B"='white'))+  theme_bw()

8.3 ggplot2气泡热图图

如果x 和 y都是离散的，把热图函数geom_tile()换成geom_point()函数，然后用表达量的值映射点的大小 同时映射颜色 也可以归为热图的一种

比如

library(readxl)dat01<-read_excel("example_data/08-heatmap/03_heatmap_example.xlsx")head(dat01)
## # A tibble: 6 x 3
##   gene_name Sample   expr_value
##   <chr>     <chr>         <dbl>
## 1 gene_1    sample_1          1
## 2 gene_1    sample_2          5
## 3 gene_1    sample_3          8
## 4 gene_1    sample_4          4
## 5 gene_1    sample_5          2
## 6 gene_1    sample_6          5
dat01$gene_name<-factor(dat01$gene_name,                        levels = c("gene_1","gene_2","gene_3",                                   "gene_4","gene_5","gene_6",                                   "gene_7","gene_8","gene_9",                                   "gene_10"))library(ggplot2)ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_point(aes(fill=expr_value,                 size=expr_value),color="blue",shape=21)

更改默认的配色和点的大小

library(readxl)dat01<-read_excel("example_data/08-heatmap/03_heatmap_example.xlsx")head(dat01)
## # A tibble: 6 x 3
##   gene_name Sample   expr_value
##   <chr>     <chr>         <dbl>
## 1 gene_1    sample_1          1
## 2 gene_1    sample_2          5
## 3 gene_1    sample_3          8
## 4 gene_1    sample_4          4
## 5 gene_1    sample_5          2
## 6 gene_1    sample_6          5
dat01$gene_name<-factor(dat01$gene_name,                        levels = c("gene_1","gene_2","gene_3",                                   "gene_4","gene_5","gene_6",                                   "gene_7","gene_8","gene_9",                                   "gene_10"))library(ggplot2)library(paletteer)ggplot(data=dat01,aes(x=gene_name,y=Sample))+  geom_point(aes(fill=expr_value,                 size=expr_value),color="blue",shape=21)+  scale_fill_paletteer_c("ggthemes::Classic Orange-White-Blue",                         direction = -1)+  scale_size_continuous(range = c(1,20),                        guide = guide_legend(override.aes = list(size = c(1,2,3,5)) ))

相关文章