How to load and split my two individual datasets when integrating datasets in Seurat?

Question

I want to do comparative analyses of my two individual single cell RNA-Seq datasets in Seurat. And I load and list my data as follows:

Control.data <- Read10X("~/Documents/...  Control_filtered_feature_bc_matrix")

Control <- CreateSeuratObject(counts  = Control.data, project = "Control_summary", min.cells = 3, min.features = 200)

Mut.data <- Read10X("~/Documents/... /Mut_filtered_feature_bc_matrix")

Mut <- CreateSeuratObject(counts  = Mut.data, project = "Mut_summary", min.cells = 3, min.features = 200)

Object<- list(Control,Mut)

names(Object) <- colnames(Object)

Then the protocol https://satijalab.org/seurat/v3.1/immune_alignment.html said, I need to use SplitObject function, but what do I split by?

Object_list <- SplitObject(Object, split.by = "?")

Answer 1

Seurat is expecting individual datasets to be normalized separately prior to data integration. In that respect lists of objects corresponding to different datasets are handy to manipulate each object/dataset individually.

In your case you do not need to do any splitting, you have an object corresponding to each of your datasets.

If you would have started with a single count matrix of all datasets then you would need to split your data based on batches/experimental condition/...