Questions about seurat-wrappers with scVelo

[ghost]

Hi,

I follow this tutorial to generate loom file with spliced and unsplied counts, followed by converting loom to Seurat object.

Here are my scripts:
ldat <- ReadVelocity(file = "path/to/loom/file")
bm <- as.Seurat(x = ldat)
saveRDS(bm, file = "path/to/rds/file")
object <- readRDS(file= "path/to/rds/file")
object[['RNA']] <- object[['unspliced']]

My questions are 1) Is it normal that the spliced/unspliced counts are much lower than the numbers I usually got from standard Seurat pipeline(without spliced/unspliced assay)? 2) Is there a way to filter out low quality cells (e.g. low nCounts_RNA, low_nFeature_RNA, high percent.mt) before I run SCTransform?

Thank you!