This is a test version of Biostars. For the public version, visit https://www.biostars.org.
How to merge multiple single cell sequencing data after processing by CellRanger software

Hi everyone, I have some questions about single cell RNA-seq analysis. I just got the output data after processing using cellranger for several patient samples, and then want to go ahead using Seurat package in R studio. But how to merge all the patient data, should I merge all the Seurat objects from each patient into one Seurat object? And how to merge that? Thanks!!!

Chen

rna-seq

Data can be aggregated using cellranger aggr see this page.

Thanks a lot for your suggestion. I'm also wondering is it possible aggregating all the data in Seurat, just saw other people using following code to merge multiple datasets and then run CCA :

sce.comb <- RunMultiCCA(list(sce1,sce2,sce3,sce4), 
                       add.cell.ids=c("PBMC_Pre.","PBMC_EarlyD27.","PBMC_RespD376.","PBMC_ARD614."),
                       genes.use = genes.use, 
                       num.cc = 30)

But it seems like so different from aggregating approach in cellranger. So the best way is to merge them in cellranger software? Thanks!

cellranger aggr merges together the data by normalization of reads mapped. running multi-sample CCA looks for correlation structures of gene-gene relationships and uses that to figure out how cells in one sample correlate best with those in another sample. I wouldn't say CCA is the same as a merge. If your aim is to compare samples to one another, CCA is the way to go. If your samples are expected to not have batch effects and be biological replicates or you don't care about the differences between them aggr may be more suitable.

1 answer

I can suggest you to try scRNA-Seq pipeline to analyse your 10x single cell RNA-Seq data using CSI NGS Portal.

You get individual as well as aggregated output files processed by Cell Ranger software.

Log in to answer this question.