读取表达矩阵创建SeuratObject

# R code
library(Seurat)
mat_5 <- read.table("GSM4037992_peripheral_donor_5_enriched_expression.tsv.gz")
row.names(mat_5) <- mat_5[,1]

#保存原始cluster labels
orig_cluster_labels5 <- mat_5$final_cluster_labels
mat_5 <- mat_5[,-c(1:3)]

#需要转置一下，行是细胞barcode
peripheral_donor_5 <- CreateSeuratObject(counts=t(mat_5),project="peripheral_donor_5")
peripheral_donor_5 <- AddMetaData(object = peripheral_donor_5, metadata = orig_cluster_labels5, col.name = "orig_cluster_labels")

mat_7 <- read.table("GSM4037994_peripheral_donor_7_enriched_expression.tsv.gz")
row.names(mat_7) <- mat_7[,1]
orig_cluster_labels7 <- mat_7$final_cluster_labels
mat_7 <- mat_7[,-c(1:3)]
peripheral_donor_7 <- CreateSeuratObject(counts=t(mat_7),project="peripheral_donor_7")
peripheral_donor_7 <- AddMetaData(object = peripheral_donor_7, metadata = orig_cluster_labels7, col.name = "orig_cluster_labels")

peripheral <- merge(peripheral_donor_5, 
                 y = peripheral_donor_7)

peripheral[["percent.mt"]] <- PercentageFeatureSet(peripheral, pattern = "^MT")

VlnPlot(peripheral, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3)

peripheral <- FindVariableFeatures(peripheral, selection.method = "vst", nfeatures = 2000)

all.genes <- rownames(peripheral)

#不是原始count 计数矩阵，直接Scale下面流程
peripheral <- ScaleData(peripheral, features = all.genes)
peripheral <- RunPCA(peripheral, features = VariableFeatures(object = peripheral))
ElbowPlot(peripheral)

peripheral <- RunUMAP(peripheral, dims = 1:20)

DimPlot(peripheral, reduction = "umap", label = TRUE, pt.size = 0.5) + NoLegend()

FeaturePlot(peripheral, features = c("ARHGEF3"))

# GSE124395

#GSE124395_Normalhumanliverdata.RData    

#An Rdata file containing the expression matrix of cells and genes from the normal human liver cell atlas. Please assign a variable and use the readRDS() function in R to load this data in R.
#这个作者有给RData文件，但是直接load发现打不开，作者让我们用readRDS读

library(Seurat)
count <- readRDS("GSE124395_Normalhumanliverdata.RData")
#dim(count)
#33941 12622
GSE124395 <- CreateSeuratObject(counts=count,project="GSE124395")
cluster <- read.table("GSE124395_clusterpartition.txt.gz")
#GSE124395部分细胞不在cluster中，应该是作者进行了过滤

#AddMetaData是可以匹配行名添加值的
GSE124395 <- AddMetaData(object = GSE124395 , metadata = cluster , col.name = "orig_cluster_labels")

#转成anndata格式
SaveH5Seurat(GSE124395, filename = "GSE124395.h5Seurat")
Convert("GSE124395.h5Seurat", dest = "h5ad")