BUSTIN

BUSTIN (BUlk and Single-cell Transcriptome data INtegration) is an R package, which identifies cell subpopulations that primarily express the genes of your interest, e.g. the drug resistant genes, by integrating bulk and single-cell transcriptome data.

Installation

BiocManager::install(c("Seurat","DESeq2","limma","cluster"))
devtools::install_github("zhangli-tools/BUSTIN")

Usage

library(Seurat)
library(BUSTIN)

Idnetify the drug resistant genes from bulk data

resistant.genes=run.limma(bulk.data, pdata, resistant=T, padj=0.05, log2fc=0.5)

if you use the bulk RNA-seq data

resistant.genes=run.DESeq(bulk.data, pdata, resistant=T, padj=0.05, log2fc=0.5)

Build a hierachical clustering tree for resistant genes based on scRNA-seq data

hc.resistant.genes=hclust.geneset(seurat.obj = seurat.obj,maxK = 10,geneset = resistant.genes) 

Predict the phenotype-associated cells

BUSTIN.out=predict.PAC(seurat.obj = seurat.obj,k.out.list = list(hc.resistant.genes),minModuleSize = 15) 

Identify the cell subpopulations associated with phenotype (drug resistance).

cell.type=ORA.celltype(BUSTIN.out$seurat.obj,features = grep("m[0-9]+$", colnames(BUSTIN.out$seurat.obj@meta.data), value=T),group.by = "label") ## group.by is the label for the cell types.