细胞类型鉴定

细胞类类型鉴定的原则

参考:https://mp.weixin.qq.com/s/3DCdxI2rwdylYwub7Vqpaw

https://pubmed.ncbi.nlm.nih.gov/32976798/

肿瘤数据分析中，第一次划分大的细胞亚群通常是：

• immune (CD45+,PTPRC),
• epithelial/cancer (EpCAM+,EPCAM),
• stromal (CD10+,MME,fibo or CD31+,PECAM1,endo)

这三类细胞亚群构成了肿瘤微环境的复杂。绝大多数文章是抓住免疫细胞的亚群进行细分，包括淋巴系（T,B,NK细胞）和髓系（单核，树突，巨噬，粒细胞）的两大类作为第二次细分亚群。但是也有不少文章是抓住stromal 里面的fibo 和endo进行细分，并且编造生物学故事的。

绝大多数文献分类的一次细胞亚群是：上皮细胞、免疫细胞、内皮细胞和成纤维细胞。最简单的比较就是不同亚群在不同生物学分组的单细胞样品比例的差异。

细胞类型鉴定的方法

图片alt

已知cell type的biomarker

• 转录因子：影响谱系lineage分化，遗传操作可控制该细胞类型
• 表面蛋白：抗体富集该种细胞类型
• 特异表达基因或基因集组合

如何找标志基因

• 看文献收集：review文章；相关领域单细胞文章。
• 数据库
  • CellMarker
  • celltaxonomy
• 实验室传承
  • 红细胞：c('HBG1','HBG2','HBQ1','HBA1','HBA2','HBE1','HBD','HBM','HBZ','HBB')
  • 性别：c('XIST','UTY','DDX3Y','TTTY15','EIF1AY','TXLNGY','RPS4Y1','KDM5D','USP9Y’)
  • 上皮细胞:c("EPCAM","CDH1","KRT5","KRT7","KRT8","KRT18","KRT19","KRT14")
  • 免疫细胞: c("PTPRC"); T cell; NK cell; B cell; Macrophage; Mast cell; Dendritic cell etc.al
  • 成纤维细胞: c("THY1", "DCN" ,"PODN", "COL1A1","COL1A2", "COL3A1","COL6A1")
  • 内皮细胞:c("PECAM1","VWF", "ENG","CDH5")

图片alt

current.cluster.ids <- c(10,17,0,3,5,12,14,15,16,1,2,6,11,7,9,8,4,13)
new.cluster.ids <- c("CC","CC","Stem","Stem","Corticotrope","Pro.PIT1","Pro.PIT1","Pro.PIT1","Pro.PIT1","Somatotrope","Somatotrope", "Somatotrope","Somatotrope","Lactotrope", "Thyrotrope","Pre.Gonado","Gonadotrope","Gonadotrope")

图片alt

在鉴定细胞类型时，计算差异基因，差异基因进行GO注释，通过GO注释功能鉴定各种细胞类型

细胞类型特异的标志基因在不同细胞群的表达

library(Seurat)
library(tidyverse)
# https://oss.bioinfo.online/cms/image/pmbc_1683691139445.rds
pbmc <- readRDS("pmbc.rds")
DotPlot(pbmc,features = c("IL7R","CCR7","CD14","LYZ","S100A4","MS4A1","CD8A","FCGR3A","MS4A7","GNLY","NKG7","FCER1A","CST3","PPBP")) +
  theme(axis.text.x=element_text(angle=90,hjust=1)) 

Correlation-based

singleR：reference里包含cell type的信息，使用singleR基于correlation查看那个相关性最高，进而鉴定细胞类型
http://bioconductor.org/books/release/SingleRBook/

Supervised Classification-based