Upload New File

scripts/1_TCR_analysis.R

+library(scRepertoire)
+library(Seurat)
+library(openxlsx)
+library(gridExtra)
+library(data.table)
+library(dplyr)
+library(reshape2)
+library(ggrepel)
+library(dittoSeq)
+library(RColorBrewer)
+library(scales)
+library(ggalluvial)
+
+########################
+### 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)
+
+samples <- c("GG-11", "GG-18", "GG-22", "GG-23", "GG-25")
+
+vdj <- list()
+for (s in samples) { 
+    vdj[[s]] <- read.csv(file = paste(wdir, "counts", paste0(s, "_counts"), "outs",
+                                      "per_sample_outs", paste0(s, "_counts"), "vdj_t",
+                                      "filtered_contig_annotations.csv", sep = "/"), 
+                         header = T)
+    vdj[[s]]$barcode <- gsub(pattern = "-1", replacement = "", x = vdj[[s]]$barcode)
+    vdj[[s]][vdj[[s]] == "true"] <- TRUE
+    vdj[[s]][vdj[[s]] == "false"] <- FALSE
+}
+
+# Step 1. 
+combined_full <- combineTCR(vdj, ID = samples, samples = samples, cells ="T-AB")
+
+combined_full_df <- rbind.data.frame(combined_full[[names(combined_full)[1]]], combined_full[[names(combined_full)[2]]])
+for(id in names(combined_full)[3:length(names(combined_full))]){
+    combined_full_df <- rbind.data.frame(combined_full_df, combined_full[[id]])
+}
+
+Full_obj <- readRDS(paste(out_dir, "Full_DBR", "Full_DBR_final_DBR_labeled.rds", sep = "/"))
+combined <- list()
+# Creating Cellid same to expression dataset
+for(id in names(combined_full)){
+    combined_full[[id]]$cellID <- "ND"
+    for (k in 1:nrow(combined_full[[id]])) {
+        combined_full[[id]]$cellID[k] <- paste0(combined_full[[id]]$sample[k], "_", strsplit(combined_full[[id]]$barcode[k], split = "_")[[1]][3], "-1")
+    }
+    combined[[id]] <- subset(combined_full[[id]], subset = cellID %in% Cells(Full_obj)) # subsetting with only ids
+    combined[[id]] <- subset(combined[[id]], subset = !is.na(TCR2)) # removing TCR2 == NA
+    print(paste0("Sample ",id, " final cells: ", nrow(combined[[id]]) ))
+}
+
+combined_df <- rbind.data.frame(combined[[names(combined)[1]]], combined[[names(combined)[2]]])
+for(id in names(combined)[3:length(names(combined))]){
+    combined_df <- rbind.data.frame(combined_df, combined[[id]])
+}
+
+##### Aggregated combining the samples ######
+A <- combined_df %>%
+    group_by(sample, cdr3_aa2) %>%
+    mutate(ClonoFreq = n()) %>%
+    group_by(sample) %>%
+    mutate(PctClonotype = ClonoFreq / n()*100) %>%
+    mutate(ClonoRel = PctClonotype / 100)
+G <- as.data.frame(A)
+rownames(G) <- G$cellID
+
+write.xlsx(x = list(VDJ = G), file = paste(out_dir, "Full_DBR", "Full_DBR_freq_by_sample_cdr3_aa2.xlsx", sep = "/"))
+saveRDS(G, paste(out_dir, "Full_DBR", "Full_DBR_freq_by_sample_cdr3_aa2.rds", sep = "/"))
+
+Full_obj <- AddMetaData(object = Full_obj, metadata = G) 
+saveRDS(object = Full_obj, file = paste(out_dir, "Full_DBR", "Full_DBR_final_DBR_labeled.rds", sep = "/"))