Skip to content

GitLab

Commit 2fd8c621 authored by Stefano Beretta's avatar Stefano Beretta
Browse files

Upload New File

parent 185b7a65
Hide whitespace changes
Inline Side-by-side
scripts/2_SubsetAnalysis_Annoations.R 0 → 100644
View file @ 2fd8c621
library(Rcpp)
library(SeuratObject)
library(Seurat)
library(harmony)
library(ggplot2)
library(patchwork)
library(RColorBrewer)
library(DoubletFinder)
library(grid)
library(cowplot)
library(gridExtra)
library(openxlsx)
library(dplyr)
########################
### General Settings ###
########################
# Working dir
wdir <- "squadrito_livertumor2022_scrnaseq"
# Raw data dir (download from GEO)
raw_dir <- "GEO_data"
# Results dir
out_dir <- paste(wdir, "results", sep = "/")
dir.create(path = out_dir, showWarnings = F)
################################
###Processing and annotattion###
################################
# Load Full object
Full_obj <- readRDS(paste(out_dir, "Full_DBR", "Full_DBR_final_DBR_labeled.rds", sep = "/"))
reg_vars <- c("CC.Difference", "percent.mt", "nCount_RNA")
# Assign group
Full_obj@meta.data$RNA_Group <- "Undefined"
Full_obj@meta.data$RNA_Group[Full_obj@meta.data$orig.ident=="GG-11"] <- "Control"
Full_obj@meta.data$RNA_Group[Full_obj@meta.data$orig.ident=="GG-18"] <- "IFNa"
Full_obj@meta.data$RNA_Group[Full_obj@meta.data$orig.ident=="GG-22"] <- "IFNa+aCTLA4"
Full_obj@meta.data$RNA_Group[Full_obj@meta.data$orig.ident=="GG-23"] <- "IFNa"
Full_obj@meta.data$RNA_Group[Full_obj@meta.data$orig.ident=="GG-25"] <- "Control"
# Annotation Newlabels in Full_obj based on computed markers
Full_obj@meta.data$CellID <- Full_obj@assays$RNA@counts@Dimnames[[2]]
Full_obj@meta.data$Newlabels <- rep("Undefined", length(Full_obj@meta.data$SingleR_ImmGenData_labels))
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "0"] <- "B cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "1"] <- "B cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "2"] <- "APCs"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "3"] <- "APCs"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "4"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "5"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "6"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "7"] <- "APCs"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "8"] <- "APCs"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "9"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "10"] <- "APCs"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "11"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "12"] <- "APCs"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "13"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "14"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "15"] <- "B cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "16"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "17"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "18"] <- "Cancer cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "19"] <- "APCs"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "20"] <- "Neutrophils"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "21"] <- "T and NK cells" # !!!
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "22"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "23"] <- "APCs"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "24"] <- "B1a-like"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "25"] <- "APCs"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "26"] <- "APCs"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "27"] <- "Undefined"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "28"] <- "Undefined"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "29"] <- "Undefined"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "30"] <- "APCs"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "31"] <- "Undefined"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "32"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "33"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "34"] <- "Undefined"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "35"] <- "Cancer cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "36"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "37"] <- "Undefined"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "38"] <- "Undefined"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "39"] <- "Undefined"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "40"] <- "T and NK cells"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "41"] <- "Undefined"
Full_obj@meta.data$Newlabels[Full_obj@meta.data$SCT_snn_res.1.2 == "42"] <- "T and NK cells"
# Split cluster 21 into "T and NK cells" and "APCs"
# Subset cluster 21 (res.1.2) which is split between APCs and T and NK cells
Full_obj <- SetIdent(object = Full_obj, value = "SCT_snn_res.1.2")
sub_obj <- subset(Full_obj, idents = c("21"))
# Process sub_obj of cluster 21
sub_obj@meta.data$CellID <- sub_obj@assays$RNA@counts@Dimnames[[2]]
DefaultAssay(object = sub_obj) <- "RNA"
sub_obj <- NormalizeData(object = sub_obj)
sub_obj <- SCTransform(sub_obj, vars.to.regress = reg_vars, verbose = TRUE)
DefaultAssay(object = sub_obj) <- "SCT"
sub_obj@misc$analysis_params$max_pca <- 50
sub_obj <- RunPCA(object = sub_obj, npcs = sub_obj@misc$analysis_params$max_pca)
ElbowPlot(sub_obj, ndims = TNK_obj@misc$analysis_params$max_pca, reduction = "pca")
sub_obj@misc$analysis_params$num_pc <- 25
sub_obj <- RunUMAP(sub_obj, dims = 1:sub_obj@misc$analysis_params$num_pc, seed.use = 1, reduction = "pca")
sub_obj <- FindNeighbors(sub_obj, reduction = "pca", dims = 1:sub_obj@misc$analysis_params$num_pc, force.recalc = T)
sub_obj <- FindClusters(sub_obj, resolution = 1.2)
#relable clusers identified in sub_obj
FeaturePlot(sub_obj, reduction = "umap", features = c("Cd3e", "Itgam"), pt.size = 2)
sub_obj@meta.data$CellID <- sub_obj@assays$RNA@counts@Dimnames[[2]]
sub_obj@meta.data$sublabels <- rep("Undefined", length(sub_obj@meta.data$SingleR_ImmGenData_labels))
sub_obj@meta.data$sublabels[sub_obj@meta.data$SCT_snn_res.1.2 == "0"] <- "T and NK cells"
sub_obj@meta.data$sublabels[sub_obj@meta.data$SCT_snn_res.1.2 == "1"] <- "T and NK cells"
sub_obj@meta.data$sublabels[sub_obj@meta.data$SCT_snn_res.1.2 == "2"] <- "APCs"
sub_obj@meta.data$sublabels[sub_obj@meta.data$SCT_snn_res.1.2 == "3"] <- "APCs"
#extract labels from sub_obj
labels_subobj <- data.frame(ToSelect = sub_obj@meta.data$CellID,
Classification = sub_obj@meta.data$sublabels)
#reintegrate sub_obj labels into full_object
alllabels <- labels_subobj
orderDF <- data.frame(allIDs = Full_obj@meta.data$CellID, order = c(1:length(Full_obj@meta.data$CellID)))
alllabels <- merge(alllabels, orderDF, by = 1)
alllabels <- alllabels[order(alllabels$order),]
Full_obj@meta.data$Newlabels[Full_obj@meta.data$CellID%in%alllabels$ToSelect] <- alllabels$Classification
# Save final object
saveRDS(object = Full_obj, paste(out_dir, "Full_DBR", "Full_DBR_final_DBR_labeled.rds", sep = "/"))
#############################
# subcluster T and NK cells #
#############################
# Subset T and NK cells
Full_obj <- SetIdent(object = Full_obj, value = "Newlabels")
TNK_obj <- subset(Full_obj, idents = c("T and NK cells"))
# Processing of the subset T cells, (Normalization, SCTransform, PCA and UMAP)
TNK_obj@meta.data$CellID <- TNK_obj@assays$RNA@counts@Dimnames[[2]]
DefaultAssay(object = TNK_obj) <- "RNA"
TNK_obj <- NormalizeData(object = TNK_obj)
TNK_obj <- SCTransform(TNK_obj, vars.to.regress = reg_vars, verbose = TRUE)
DefaultAssay(object = TNK_obj) <- "SCT"
TNK_obj@misc$analysis_params$max_pca <- 50
TNK_obj <- RunPCA(object = TNK_obj, npcs = TNK_obj@misc$analysis_params$max_pca)
ElbowPlot(TNK_obj, ndims = TNK_obj@misc$analysis_params$max_pca, reduction = "pca")
TNK_obj@misc$analysis_params$num_pc <- 40
TNK_obj <- RunUMAP(TNK_obj, dims = 1:TNK_obj@misc$analysis_params$num_pc, seed.use = 1, reduction = "pca")
TNK_obj <- FindNeighbors(TNK_obj, reduction = "pca", dims = 1:TNK_obj@misc$analysis_params$num_pc, force.recalc = T)
TNK_obj <- FindClusters(TNK_obj, resolution = 1.2)
# Annotate Newlabels_rough into TNK_obj
TNK_obj@meta.data$Newlabels_rough <- rep("Undefined",length(TNK_obj@meta.data$SingleR_ImmGenData_labels))
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "0"] <- "NK"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "1"] <- "NKT"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "2"] <- "CD4 T cells"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "3"] <- "CD8 T cells"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "4"] <- "CD4 T cells"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "5"] <- "CD8 T cells"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "6"] <- "CD4 T cells"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "7"] <- "CD4 T cells"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "8"] <- "ILCs"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "9"] <- "CD8 T cells"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "10"] <- "CD8 T cells"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "11"] <- "CD8 T cells"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "12"] <- "CD8 T cells"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "13"] <- "Undefined"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "14"] <- "T prol"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "15"] <- "Undefined"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "16"] <- "Undefined"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "17"] <- "gd T cells"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "18"] <- "Undefined"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "19"] <- "Undefined"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "20"] <- "Undefined"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "21"] <- "Undefined"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "22"] <- "Undefined"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "23"] <- "Undefined"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "24"] <- "Undefined"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "25"] <- "Undefined"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "26"] <- "Undefined"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "27"] <- "Undefined"
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$SCT_snn_res.1.2 == "28"] <- "Undefined"
# Detailed annotation NKT cells
# Subset NK and NKT cells to TNKonly_obj
TNK_obj <- SetIdent(object = TNK_obj, value = "Newlabels_rough")
TNKonly_obj <- subset(TNK_obj, idents = c("NKT"))
# Process TNKonly_obj
DefaultAssay(object = TNKonly_obj) <- "SCT"
TNKonly_obj@misc$analysis_params$max_pca <- 50
TNKonly_obj <- RunPCA(object = TNKonly_obj, npcs = TNKonly_obj@misc$analysis_params$max_pca)
ElbowPlot(TNKonly_obj, ndims = TNKonly_obj@misc$analysis_params$max_pca, reduction = "pca")
TNKonly_obj@misc$analysis_params$num_pc <- 30
TNKonly_obj <- RunUMAP(TNKonly_obj, dims = 1:TNKonly_obj@misc$analysis_params$num_pc, seed.use = 1, reduction = "pca")
TNKonly_obj <- FindNeighbors(TNKonly_obj ,reduction = "pca", dims = 1:TNKonly_obj@misc$analysis_params$num_pc, force.recalc = T)
TNKonly_obj <- FindClusters(TNKonly_obj , resolution = 1.4)
# Detailed annotation Newlabels_detailed of TNKonly_obj
TNKonly_obj@meta.data$Newlabels_detailed <- rep("Undefined",length(TNKonly_obj@meta.data$SingleR_ImmGenData_labels))
TNKonly_obj@meta.data$Newlabels_detailed[TNKonly_obj@meta.data$SCT_snn_res.1.4 == "0"] <- "MAIT"
TNKonly_obj@meta.data$Newlabels_detailed[TNKonly_obj@meta.data$SCT_snn_res.1.4 == "1"] <- "MAIT"
TNKonly_obj@meta.data$Newlabels_detailed[TNKonly_obj@meta.data$SCT_snn_res.1.4 == "2"] <- "MAIT"
TNKonly_obj@meta.data$Newlabels_detailed[TNKonly_obj@meta.data$SCT_snn_res.1.4 == "3"] <- "MAIT"
TNKonly_obj@meta.data$Newlabels_detailed[TNKonly_obj@meta.data$SCT_snn_res.1.4 == "4"] <- "MAIT"
TNKonly_obj@meta.data$Newlabels_detailed[TNKonly_obj@meta.data$SCT_snn_res.1.4 == "5"] <- "gd NKT"
TNKonly_obj@meta.data$Newlabels_detailed[TNKonly_obj@meta.data$SCT_snn_res.1.4 == "6"] <- "MAIT"
TNKonly_obj@meta.data$Newlabels_detailed[TNKonly_obj@meta.data$SCT_snn_res.1.4 == "7"] <- "MAIT"
TNKonly_obj@meta.data$Newlabels_detailed[TNKonly_obj@meta.data$SCT_snn_res.1.4 == "8"] <- "MAIT"
TNKonly_obj@meta.data$Newlabels_detailed[TNKonly_obj@meta.data$SCT_snn_res.1.4 == "9"] <- "MAIT"
TNKonly_obj@meta.data$Newlabels_detailed[TNKonly_obj@meta.data$SCT_snn_res.1.4 == "10"] <- "MAIT"
# Extract labels from NKT cells
labels_NKT <- data.frame(ToSelect = TNKonly_obj@meta.data$CellID,
Classification = TNKonly_obj@meta.data$Newlabels_detailed)
# Integrate Newlabels_detailed annotation from TNKonly_obj into TNK_obj
TNK_obj@meta.data$Newlabels_detailed <- TNK_obj@meta.data$Newlabels_rough
alllabels <- labels_NKT
TNK_obj@meta.data$Newlabels_detailed <- TNK_obj@meta.data$Newlabels_rough
orderDF <- data.frame(allIDs = TNK_obj@meta.data$CellID,
order=c(1:length(TNK_obj@meta.data$CellID)))
alllabels <- merge(alllabels, orderDF, by = 1)
alllabels <- alllabels[order(alllabels$order),]
TNK_obj@meta.data$Newlabels_detailed[TNK_obj@meta.data$CellID%in%alllabels$ToSelect] <- alllabels$Classification
## Detailed annotation CD4 T cells ##
# Generate CD4 T cell subset
TNK_obj <- SetIdent(object = TNK_obj, value = "Newlabels_rough")
CD4_obj <- subset(TNK_obj, idents = c("CD4 T cells"))
# Process CD4_obj
DefaultAssay(object = CD4_obj) <- "SCT"
CD4_obj@misc$analysis_params$max_pca <- 50
CD4_obj <- RunPCA(object = CD4_obj, npcs = CD4_obj@misc$analysis_params$max_pca)
ElbowPlot(CD4_obj, ndims = CD4_obj@misc$analysis_params$max_pca, reduction = "pca")
CD4_obj@misc$analysis_params$num_pc <- 35
CD4_obj <- RunUMAP(CD4_obj, dims = 1:CD4_obj@misc$analysis_params$num_pc, seed.use = 1, reduction = "pca")
CD4_obj <- FindNeighbors(CD4_obj, reduction = "pca", dims = 1:CD4_obj@misc$analysis_params$num_pc, force.recalc = T)
CD4_obj <- FindClusters(CD4_obj, resolution = 0.5)
# Detailed annotation Newlabels_detailed of CD4_obj
CD4_obj@meta.data$Newlabels_detailed <- rep("Undefined",length(CD8_obj@meta.data$SingleR_ImmGenData_labels))
CD4_obj@meta.data$Newlabels_detailed[CD4_obj@meta.data$SCT_snn_res.0.5 == "0"] <- "CD4 Treg cells"
CD4_obj@meta.data$Newlabels_detailed[CD4_obj@meta.data$SCT_snn_res.0.5 == "1"] <- "CD4 Teff cells"
CD4_obj@meta.data$Newlabels_detailed[CD4_obj@meta.data$SCT_snn_res.0.5 == "2"] <- "CD4 Tex cells"
CD4_obj@meta.data$Newlabels_detailed[CD4_obj@meta.data$SCT_snn_res.0.5 == "3"] <- "CD4 Tsm-like cells"
CD4_obj@meta.data$Newlabels_detailed[CD4_obj@meta.data$SCT_snn_res.0.5 == "4"] <- "CD8 T cells"
# Extract labels from CD4 T cells
labels_CD4 <- data.frame(ToSelect = CD4_obj@meta.data$CellID,
Classification = CD4_obj@meta.data$Newlabels_detailed)
# Integrate Newlabels_detailed annotation from CD4_obj into TNK_obj
alllabels <- labels_CD4
orderDF <- data.frame(allIDs = TNK_obj@meta.data$CellID, order = c(1:length(TNK_obj@meta.data$CellID)))
alllabels <- merge(alllabels, orderDF, by = 1)
alllabels <- alllabels[order(alllabels$order),]
TNK_obj@meta.data$Newlabels_detailed[TNK_obj@meta.data$CellID%in%alllabels$ToSelect] <- alllabels$Classification
# Annotate newly identified CD8 T cells in CD4_obj as CD8 T cells in Newlabels_rough slot in TNK_obj
TNK_obj@meta.data$Newlabels_rough[TNK_obj@meta.data$Newlabels_detailed=="CD8 T cells"] <- "CD8 T cells"
## Detailed annotation CD8 T cells ##
# Generate CD8 T cell subset
TNK_obj <- SetIdent(object = TNK_obj, value = "Newlabels_rough")
CD8_obj <- subset(TNK_obj, idents = c("CD8 T cells"))
# Process CD8_obj
DefaultAssay(object = CD8_obj) <- "SCT"
CD8_obj@misc$analysis_params$max_pca <- 50
CD8_obj <- RunPCA(object = CD8_obj, npcs = CD8_obj@misc$analysis_params$max_pca)
ElbowPlot(CD8_obj, ndims = CD8_obj@misc$analysis_params$max_pca, reduction = "pca")
CD8_obj@misc$analysis_params$num_pc <- 35
CD8_obj <- RunUMAP(CD8_obj, dims = 1:CD8_obj@misc$analysis_params$num_pc, seed.use = 1, reduction = "pca")
CD8_obj <- FindNeighbors(CD8_obj, reduction = "pca", dims = 1:CD8_obj@misc$analysis_params$num_pc, force.recalc = T)
CD8_obj <- FindClusters(CD8_obj, resolution = 0.9)
# Detailed annotation Newlabels_detailed of CD8_obj
CD8_obj@meta.data$Newlabels_detailed <- rep("Undefined",length(CD8_obj@meta.data$SingleR_ImmGenData_labels))
CD8_obj@meta.data$Newlabels_detailed[CD8_obj@meta.data$SCT_snn_res.0.9 == "0"] <- "CD8 Teff_ex cells"
CD8_obj@meta.data$Newlabels_detailed[CD8_obj@meta.data$SCT_snn_res.0.9 == "1"] <- "CD8 Tm cells"
CD8_obj@meta.data$Newlabels_detailed[CD8_obj@meta.data$SCT_snn_res.0.9 == "2"] <- "CD8 Tsm-like cells"
CD8_obj@meta.data$Newlabels_detailed[CD8_obj@meta.data$SCT_snn_res.0.9 == "3"] <- "CD8 Teff_ex cells"
CD8_obj@meta.data$Newlabels_detailed[CD8_obj@meta.data$SCT_snn_res.0.9 == "4"] <- "CD8 Tsm-like cells"
CD8_obj@meta.data$Newlabels_detailed[CD8_obj@meta.data$SCT_snn_res.0.9 == "5"] <- "CD8 Teff_ex cells"
CD8_obj@meta.data$Newlabels_detailed[CD8_obj@meta.data$SCT_snn_res.0.9 == "6"] <- "Undefined" #CD8a low
CD8_obj@meta.data$Newlabels_detailed[CD8_obj@meta.data$SCT_snn_res.0.9 == "7"] <- "CD8 Teff_ex cells"
CD8_obj@meta.data$Newlabels_detailed[CD8_obj@meta.data$SCT_snn_res.0.9 == "8"] <- "CD8 Teff_ex cells"
CD8_obj@meta.data$Newlabels_detailed[CD8_obj@meta.data$SCT_snn_res.0.9 == "9"] <- "CD8 Teff_ex cells"
CD8_obj@meta.data$Newlabels_detailed[CD8_obj@meta.data$SCT_snn_res.0.9 == "10"] <- "Undefined"
CD8_obj@meta.data$Newlabels_detailed[CD8_obj@meta.data$SCT_snn_res.0.9 == "11"] <- "Undefined"
# Extract labels from CD8 T cells
labels_CD8 <- data.frame(ToSelect = CD8_obj@meta.data$CellID,
Classification = CD8_obj@meta.data$Newlabels_detailed)
# Integrate Newlabels_detailed annotation from CD8_obj into TNK_obj
alllabels <- labels_CD8
orderDF <- data.frame(allIDs = TNK_obj@meta.data$CellID, order = c(1:length(TNK_obj@meta.data$CellID)))
alllabels <- merge(alllabels, orderDF, by = 1)
alllabels <- alllabels[order(alllabels$order),]
TNK_obj@meta.data$Newlabels_detailed[TNK_obj@meta.data$CellID%in%alllabels$ToSelect] <- alllabels$Classification
## Remove undefined cells
TNK_obj <- SetIdent(object = TNK_obj, value = "Newlabels_detailed")
TNK_obj <- subset(TNK_obj, idents = c("Undefined"), invert = T)
# Add analysis of clonotype -> cathegorize T cells into expansion levels
a1 <- 25
a2 <- 15
a3 <- 5
a4 <- 1
TNK_obj@meta.data$cloneType <- "NA"
TNK_obj@meta.data$cloneType[TNK_obj@meta.data$ClonoFreq2 > a1 ] <- paste0("Hyperexpanded (x>", a1, ")")
TNK_obj@meta.data$cloneType[TNK_obj@meta.data$ClonoFreq2 > a2 & TNK_obj@meta.data$ClonoFreq2 <= a1] <- paste0("Large (", a2, "< x <=", a1, ")")
TNK_obj@meta.data$cloneType[TNK_obj@meta.data$ClonoFreq2 > a3 & TNK_obj@meta.data$ClonoFreq2 <= a2] <- paste0("Medium (", a3, "< x <=", a2, ")")
TNK_obj@meta.data$cloneType[TNK_obj@meta.data$ClonoFreq2 > a4 & TNK_obj@meta.data$ClonoFreq2 <= a3] <- paste0("Small(", a4, "< x <=", a3, ")")
TNK_obj@meta.data$cloneType[TNK_obj@meta.data$ClonoFreq2 == a4 ] <- paste0("Unique(x=", a4, ")")
# Save TNK_obj
dir.create(paste(out_dir, "TNK", sep = "/"))
saveRDS(TNK_obj, file = paste(out_dir, "TNK", "TNK_DBR_final.rds", sep = "/"))
#####################
## Subcluster APCs ##
#####################
Full_obj <- SetIdent(object = Full_obj, value = "Newlabels")
APC_obj <- subset(Full_obj, idents = c("APCs"))
# Processing of the APC_obj, (Normalization, SCTransform, PCA and UMAP)
APC_obj@meta.data$CellID <- APC_obj@assays$RNA@counts@Dimnames[[2]]
DefaultAssay(object = APC_obj) <- "RNA"
APC_obj <- NormalizeData(object = APC_obj)
APC_obj <- SCTransform(APC_obj, vars.to.regress = reg_vars, verbose = TRUE)
DefaultAssay(object = APC_obj) <- "SCT"
APC_obj@misc$analysis_params$max_pca <- 50
APC_obj <- RunPCA(object = APC_obj, npcs = APC_obj@misc$analysis_params$max_pca)
ElbowPlot(APC_obj, ndims = APC_obj@misc$analysis_params$max_pca, reduction = "pca")
APC_obj@misc$analysis_params$num_pc <- 35
APC_obj <- RunUMAP(APC_obj, dims = 1:APC_obj@misc$analysis_params$num_pc, seed.use = 1, reduction = "pca")
APC_obj <- FindNeighbors(APC_obj, reduction = "pca", dims = 1:APC_obj@misc$analysis_params$num_pc, force.recalc = T)
APC_obj <- FindClusters(APC_obj, resolution = 1.2)
# Annotate Newlabels_rough on macrophages
APC_obj@meta.data$Newlabels_rough <- rep("Undefined",length(APC_obj@meta.data$SingleR_ImmGenData_labels))
APC_obj@meta.data$CellID <- APC_obj@assays$RNA@counts@Dimnames[[2]]
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "0"] <- "Monocyte & Macrophages"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "1"] <- "Monocyte & Macrophages"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "2"] <- "Monocyte & Macrophages"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "3"] <- "Mo DCs"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "4"] <- "Monocyte & Macrophages"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "5"] <- "Monocyte & Macrophages"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "6"] <- "Monocyte & Macrophages"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "7"] <- "Monocyte & Macrophages"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "8"] <- "Monocyte & Macrophages"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "9"] <- "DCs"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "10"] <- "DCs"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "11"] <- "Mo DCs"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "12"] <- "Undefined"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "13"] <- "DCs"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "14"] <- "Monocyte & Macrophages"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "15"] <- "DCs"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "16"] <- "Monocyte & Macrophages"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "17"] <- "Mo DCs"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "18"] <- "Undefined"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "19"] <- "Undefined"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "20"] <- "Undefined"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "21"] <- "Undefined"
APC_obj@meta.data$Newlabels_rough[APC_obj@meta.data$SCT_snn_res.1.2 == "22"] <- "Undefined"
# Annotate Newlabels_detailed on Macrophages
APC_obj@meta.data$Newlabels_detailed <- rep("Undefined",length(APC_obj@meta.data$SingleR_ImmGenData_labels))
APC_obj@meta.data$CellID <- APC_obj@assays$RNA@counts@Dimnames[[2]]
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "0"] <- "TAMs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "1"] <- "KCs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "2"] <- "IFNa TAMs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "3"] <- "Mo DCs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "4"] <- "IFNa TAMs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "5"] <- "IFNa TAMs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "6"] <- "IFNa TAMs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "7"] <- "TAMs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "8"] <- "IFNa TAMs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "9"] <- "CDPs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "10"] <- "cDC1"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "11"] <- "Mo DCs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "12"] <- "Undefined"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "13"] <- "pDCs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "14"] <- "Monocytes"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "15"] <- "Ccr7 DCs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "16"] <- "KCs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "17"] <- "Mo DCs"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "18"] <- "Undefined"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "19"] <- "Undefined"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "20"] <- "Undefined"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "21"] <- "Undefined"
APC_obj@meta.data$Newlabels_detailed[APC_obj@meta.data$SCT_snn_res.1.2 == "22"] <- "Undefined"
# Remove undefined cells
APC_obj <- SetIdent(object = APC_obj, value = "Newlabels_detailed")
APC_obj <- subset(APC_obj, idents = c("Undefined"), invert = T)
# save APCs object
dir.create(paste(out_dir, "APCs", sep = "/"))
saveRDS(APC_obj, file = paste(out_dir, "APCs", "APCs_DBR_final.rds", sep = "/"))
## Reintegrate all labeles into Full object
# Reintegrate lables from TNK_obj into Full_obj
# Newlables_rough
labels_TNK_rough <- data.frame(ToSelect = TNK_obj@meta.data$CellID,
Classification = TNK_obj@meta.data$Newlabels_rough)
alllabels <- labels_TNK_rough
orderDF <- data.frame(allIDs = TNK_obj@meta.data$CellID, order = c(1:length(TNK_obj@meta.data$CellID)))
alllabels <- merge(alllabels, orderDF, by = 1)
alllabels <- alllabels[order(alllabels$order),]
Full_obj@meta.data$Newlabels_rough<-Full_obj@meta.data$Newlabels
Full_obj@meta.data$Newlabels_rough[Full_obj@meta.data$CellID%in%alllabels$ToSelect] <- alllabels$Classification
DimPlot(object = Full_obj, reduction = "umap", group.by = "Newlabels_rough", pt.size = 1, label = T, label.size = 10)
# Newlabels_detailed
labels_TNK_detailed <- data.frame(ToSelect = TNK_obj@meta.data$CellID,
Classification = TNK_obj@meta.data$Newlabels_detailed)
alllabels <- labels_TNK_detailed
orderDF <- data.frame(allIDs = TNK_obj@meta.data$CellID, order = c(1:length(TNK_obj@meta.data$CellID)))
alllabels <- merge(alllabels, orderDF, by = 1)
alllabels <- alllabels[order(alllabels$order),]
Full_obj@meta.data$Newlabels_detailed<-Full_obj@meta.data$Newlabels
Full_obj@meta.data$Newlabels_detailed[Full_obj@meta.data$CellID%in%alllabels$ToSelect] <- alllabels$Classification
DimPlot(object = Full_obj, reduction = "umap", group.by = "Newlabels_detailed", pt.size = 1, label = T, label.size = 4)
# Reintegrate lables from APC_obj into Full_obj
# Newlables_rough
labels_APC_rough <- data.frame(ToSelect = APC_obj@meta.data$CellID,
Classification = APC_obj@meta.data$Newlabels_rough)
alllabels <- labels_APC_rough
orderDF <- data.frame(allIDs = APC_obj@meta.data$CellID, order = c(1:length(APC_obj@meta.data$CellID)))
alllabels <- merge(alllabels, orderDF, by = 1)
alllabels <- alllabels[order(alllabels$order),]
Full_obj@meta.data$Newlabels_rough[Full_obj@meta.data$CellID%in%alllabels$ToSelect] <- alllabels$Classification
DimPlot(object = Full_obj, reduction = "umap", group.by = "Newlabels_rough", pt.size = 1, label = T, label.size = 10)
# Newlabels_detailed
labels_APC_detailed <- data.frame(ToSelect = APC_obj@meta.data$CellID,
Classification = APC_obj@meta.data$Newlabels_detailed)
alllabels <- labels_APC_detailed
orderDF <- data.frame(allIDs = APC_obj@meta.data$CellID, order = c(1:length(APC_obj@meta.data$CellID)))
alllabels <- merge(alllabels, orderDF, by = 1)
alllabels <- alllabels[order(alllabels$order),]
Full_obj@meta.data$Newlabels_detailed[Full_obj@meta.data$CellID%in%alllabels$ToSelect] <- alllabels$Classification
DimPlot(object = Full_obj, reduction = "umap", group.by = "Newlabels_detailed", pt.size = 1, label = T, label.size = 4)
# Remove undefined cells (cells annotated as undefined or cells that were removed in TNK_obj and APC_obj analysis)
Full_obj <- SetIdent(object = Full_obj, value = "Newlabels_detailed")
Full_obj <- subset(Full_obj, idents = c("Undefined","T and NK cells"), invert = T)
# Save Full_obj
saveRDS(object = Full_obj, paste(out_dir, "Full_DBR", "Full_DBR_final_DBR_labeled.rds", sep = "/"))
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment