SCTransform taking hours with very large memory demand

satijalab / sctransform

R package for modeling single cell UMI expression data using regularized negative binomial regression

GNU General Public License v3.0

206 stars 33 forks source link

SCTransform taking hours with very large memory demand #57

Open liqianma7 opened 4 years ago

liqianma7 commented 4 years ago

I have a very large dataset (Seurat object ~8gb) and sctransform takes quite long time and eventually failed due to memory issue (RAM=80gb). In this case, would you recommend doing sctransform only on variable features? Will this affect DEG analysis downstream?

Thank you very much for your help!

ChristophH commented 4 years ago

What are the dimensions of your input matrix? What are the exact commands you use? It's hard to pinpoint the issue without that information. If you are using Seurat::SCTransform you might want to set conserve.memory = TRUE