单细胞专题（2） | 亚群细化分析并寻找感兴趣的小亚群

bioyigene

已于 2023-07-24 11:13:28 修改

阅读量5.4k

点赞数 1

文章标签：经验分享

于 2023-04-10 13:25:49 首次发布

本文链接：https://blog.csdn.net/bioyigene/article/details/130057654

版权

通常情况下，单细胞转录组拿到亚群后会进行更细致的分群，或者看不同样本不同组别的内部的细胞亚群的比例变化。

这就是个性化分析阶段，这个阶段取决于自己的单细胞转录组项目课题设计情况，我们了解到的各式各样的分析点，并不是通用的。比如如果要比较细胞亚群比例，就必须要有多个样本，如果是单个样本，其分析的内容及方法也会不尽相同，类似的问题以及涉及到的分析很多。

今天小编带大家来学习下“单细胞个性化分析之细胞亚群继续分群”，相信大多数单细胞转录组的小伙伴都能用到。

细胞亚群进一步划分原则

理论上细胞亚群是可以无限划分的，因为世界上没有两个一模一样的细胞，关键是要把握一个度，什么样的差异可以判定为不同细胞亚群，什么样的差异是可以容忍的细胞类群内部异质性。有一个策略就是找出主要因素和次要因素。主要因素划分为主要亚群，比如外周血里面的T,B细胞当然是不同亚群，但是T细胞里面还可以继续划分：CD4或者CD8的T细胞，甚至继续划分，如下图所示：

亚群细化分析实操

加载R包

{
library(Seurat)
library(dplyr)
library(reticulate)
library(sctransform)
library(cowplot)
library(ggplot2)
library(viridis)
library(tidyr)
library(magrittr)
library(reshape2)
library(readxl)
library(stringr)
library(cowplot)
library(scales)
library(readr)
library(progeny)
library(gplots)
library(tibble)
library(grid)
library(rlang)

theme_set(theme_cowplot())
use_colors <- c(
  Alveolar_Macrophages = "#6bAEd6",
  `Monocyte-derived macrophages`= "#fff500",
  Monocytes= "#FA9FB5",
  `Myeloid dendritic cells`= "#DD3497",
  Plasmacytoid_dendritic_cells= "#7A0177",
  `Conventional_T_cells`= "#c2e699",
  `Regulatory T cells`= "#006837",
  `CD8+_T_cells`= "#bcbddc",
  NK_cells= "#4a1486",
  B_cells= "#969696",
  Plasma_cells= "#636363")
}
imm_anno <- readRDS("imm_anno.RDS")
imm_anno@meta.data$cell_type_imm <-
  ordered(
    imm_anno@meta.data$cell_type_imm, levels = c("Alveolar_Macrophages",  
                                       "Monocyte-derived macrophages", 
                                       "Monocytes","Myeloid dendritic cells", 
                                       "Plasmacytoid_dendritic_cells",
                                       "Conventional_T_cells", 
                                       "Regulatory T cells", 
                                       "CD8+_T_cells", 
                                       "NK_cells", 
                                       "B_cells", 
                                       "Plasma_cells")
           )
DimPlot(imm_anno, reduction = 'umap', split.by = 'tissue_type', label = T, cols = use_colors)
##产生聚类点图

细化分析的开端是发现（感觉）某群细胞有多个亚群，如下面的图发现单核来源巨噬有多群。

对巨噬细胞分析M1,M2表型通路

### 实际上同一群细胞内部小亚群比较更有意思
###导入M1
m1m2_pws <- read_lines("./data/CLASSICAL_M1_VS_ALTERNATIVE_M2_MACROPHAGE_UP.gmt") %>%
  lapply(str_split, "\\t") %>% 
  unlist(recursive = F) %>% 
  lapply(function(x) setNames(list(x[-c(1:2)]), x[1])) %>% 
  unlist(recursive = F)

### 导入M2
m1m2_pws <- append(m1m2_pws, read_lines("./data/CLASSICAL_M1_VS_ALTERNATIVE_M2_MACROPHAGE_DN.gmt") %>%
                     lapply(str_split, "\\t") %>% 
                     unlist(recursive = F) %>% 
                     lapply(function(x) setNames(list(x[-c(1:2)]), x[1])) %>% 
                     unlist(recursive = F))

### 运用addmodulescore函数，加入富集分数到metadata中
imm_anno <- AddModuleScore(object = imm_anno, features = m1m2_pws, name = c("m1up", "m1dn"), nbin = 12)### 提出单核来源巨噬来分析macrophage（mdm）
scRNA_mdm <- subset(imm_anno, subset = cell_type_imm %in% c("Monocyte-derived macrophages"))
scRNA_mdm=ScaleData(scRNA_mdm)

### 重新降维聚类
scRNA_mdm <- RunPCA(scRNA_mdm)
ElbowPlot(scRNA_mdm,  ndims = 50)

library(ggplot2)
scRNA_mdm<- RunUMAP(scRNA_mdm, dims = 1:20)
scRNA_mdm <- FindNeighbors(scRNA_mdm, dims = 1:20)
for (i in c(0.01,0.2, 0.3, 0.4, 0.5, 1)) {
  scRNA_mdm <- FindClusters(scRNA_mdm, resolution = i)
  print(DimPlot(scRNA_mdm, reduction = "umap") + labs(title = paste0("resolution: ", i)))
}
### cluster数自己决定，不要太多即可,否则相似的群多
table(scRNA_mdm$SCT_snn_res.0.2)
Idents(scRNA_mdm) <-scRNA_mdm@meta.data$SCT_snn_res.0.2

DimPlot(scRNA_mdm,reduction = 'umap',group.by = 'SCT_snn_res.0.2',label = T,split.by = 'tissue_type')
ggsave2(filename = 'Macrophage亚群.pdf',path = './result_3v3/analysis',width = 8,height = 4)

### 不同mdm小亚群比较通路
VlnPlot(scRNA_mdm, features = c("m1up1", "m1dn2"), 
        group.by = "SCT_snn_res.0.2", pt.size = 0, 
        cols = use_colors)
### 第四群有一个显著的M1活化
ggsave2(filename = 'M1M2巨噬细胞细分.pdf',path = './result_3v3/analysis/',width = 8,height =4)

### 单核细胞来源的巨噬细胞间具有较大异质性，所以考虑再分别注释

mdm_markers <- FindAllMarkers(scRNA_mdm, only.pos = T, min.pct = 0.25, min.diff.pct = 0.25)

# group_by和top_n组合
top_mdm_markers <- mdm_markers %>% group_by(cluster) %>% top_n(10, wt = avg_log2FC)

DoHeatmap(scRNA_mdm,features=top_mdm_markers$gene,group.by = 'SCT_snn_res.0.2')+scale_fill_viridis()
ggsave2(filename = '热图细分.pdf',path = './result_3v3/analysis/',width = 10,height =9)

#### 第4群CCL5+，其实还有CD8A+，大家认为，这是一群新的巨噬，还是由于细胞污染呢~
FeaturePlot(scRNA_mdm,features = 'CCL5',cols = viridis(10),label = T)
ggsave2(filename = 'CCL5_mdm.pdf',path = './result_3v3/analysis/',width = 5,height =4.5)

### 据此我们可以继续定义细胞亚群，例如cluster1为C1Q+的巨噬细胞！