What is the right way to merge multiple patient samples in scrna-seq?

Question

[cogen859]

Hi all,

I have a single cell dataset which comprises of cell expression files from multiple patients. I am curious what is the right way to merge these files. I know there is a integration vignette in Seurat but I felt its more like integrating samples that are different in terms of condition/therapy etc. In my case all the samples are tumor samples. Just importing expression files and merging using "merge" function in Seurat is enough. Or do I need to perform integration using CCA/RPCA?

Thank you for your guidance.

[ATpoint]

If there are multiple samples then there is quite a chance that batch effects will influence any clustering, so it is usually standard to do the integration. Only one way to find out though. Merge them, cluster them, and then color the plot by sample and see whether you find evidence that entire clusters are separate due to batch/sample rather than biology. I would bet money though that integration will be necessary (it almost always is from what I've seen and read so far).