Entering edit mode
19 months ago
Nodilan
▴
10
Hello, I have filtred seurat object for different replicates ( control and patients ) , my objective is to determine differential expressed genes between 2 conditions my question: is it compulsory to perform SCTTransform before data integration ? if it's the case why ? I'm using this tutorial from seurat as a reference and they didn't so scttransform :/ https://satijalab.org/seurat/archive/v3.0/immune_alignment.html
No, there are many methods available for integration. The vignette you link is a version before SCtransform became the default, simple as that.
thank you for responding, In this I can use the old version because when I apply sctransform I get memory problem when I integrate data :/
You can also use something like harmony which uses less memory and accepts Seurat objects.
Note that if you have biological replicates for each condition it's better to do pseudobulk differential expression after integration. See edgeR (section 4.10) or OSCA for more information. Because of this integration only matters insofar as getting accurate clustering and cell type annotation.