大漂亮版本的单细胞小提琴图

0.简介

seurat自带的小提琴图，在基因数量多时变得不易读。看到一个漂亮的版本。

比较如下：

1.数据

rm(list = ls())
load("obj.Rdata")
load("markers.Rdata")
ls()

## [1] "pbmc"         "pbmc.markers"

class(pbmc)

## [1] "Seurat"
## attr(,"package")
## [1] "SeuratObject"

head(pbmc.markers)

##        p_val avg_log2FC pct.1 pct.2 p_val_adj cluster   gene
## S100A9     0   4.165490 0.997 0.092         0       0 S100A9
## S100A8     0   3.476109 0.956 0.048         0       0 S100A8
## PPIF       0   3.251813 0.989 0.128         0       0   PPIF
## CD300E     0   3.002149 0.972 0.082         0       0 CD300E
## S100A6     0   2.877628 1.000 0.565         0       0 S100A6
## TYROBP     0   2.818819 1.000 0.209         0       0 TYROBP

pbmc数据是seurat标准流程findcluster得到的，pbmc.markers是findallmarkers得到的结果。

2.选择基因

library(tidyverse)
g = pbmc.markers %>% 
  group_by(cluster) %>% 
  top_n(n = 2, wt = avg_log2FC) %>% 
  pull(gene)

这个数据总共13个簇，选出了26个基因。

3.准备画图数据并画图

自定义的图，不能使用seurat自带的函数啦。要自己搞宽变长，然后用ggplot2画。

vln.df <- pbmc@assays$RNA@scale.data %>%
  t() %>%
  as.data.frame()%>%
  select(g) %>% 
  rownames_to_column("CB") %>% 
  mutate(cluster = pbmc$RNA_snn_res.0.5)%>%
  pivot_longer(cols = 2:(ncol(.)-1),
               names_to = "gene",
               values_to = "exp") %>% 
  mutate(gene = factor(gene,levels = g))
head(vln.df)

## # A tibble: 6 × 4
##   CB               cluster gene         exp
##   <chr>            <fct>   <fct>      <dbl>
## 1 AAACCCAAGTAGGGTC 8       S100A9    -0.640
## 2 AAACCCAAGTAGGGTC 8       S100A8    -0.571
## 3 AAACCCAAGTAGGGTC 8       GIMAP7    -0.329
## 4 AAACCCAAGTAGGGTC 8       CD3G      -0.704
## 5 AAACCCAAGTAGGGTC 8       LINC02446 -0.348
## 6 AAACCCAAGTAGGGTC 8       IL7R      -0.390

# 自定义颜色
library(paletteer)
my_color = paletteer_d(`"ggsci::default_nejm"`)
my_color = colorRampPalette(my_color)(length(unique(vln.df$cluster)))
# 画图
p1 <- ggplot(vln.df,aes(cluster,exp),color=factor(cluster))+
  geom_violin(aes(fill=cluster),scale = "width")+
  scale_fill_manual(values = my_color)+
  facet_grid(gene~.,scales = "free_y")+
  scale_y_continuous(expand = c(0,0))+
  theme_bw()+
  theme(
    panel.grid = element_blank(),
    axis.title.x.bottom = element_blank(),
    axis.ticks.x.bottom = element_blank(),
    axis.text.x.bottom = element_text(hjust = 1,vjust = NULL,color = "black",size = 14),
    axis.title.y.left = element_blank(),
    axis.ticks.y.left = element_blank(),
    axis.text.y.left = element_blank(),
    legend.position = "none",
    panel.spacing.y = unit(0, "cm"),
    strip.text.y = element_text(angle=0,size = 14,hjust = 0),
    strip.background.y = element_blank()
  )
p1

参考：https://blog.csdn.net/qq_38774801/article/details/130791574

最后编辑于：2023.06.23 08:33:09