Findallmarkers on seurat object after SCTransform and integration

[ssukumaran2]

Hi,
I have a question about using FindAllMarkers on a seurat object generated by integration of six biological replicates after SCTransform v2. I have done PrepSCTFindMarkers on this object. Is there anything else that need to be done before running FindAllMarkers? DE analysis has not worked well using this method- I found far fewer marker genes compared to a DE analysis on the RNA assay, and many that I know should be present are missing.

Please see my code below:
obj[["RNA"]] <- split(obj[["RNA"]], f = obj$orig.ident)
obj<-SCTransform(obj, vst.flavor=‘v2', vars.to.regress = 'mitoRatio',verbose = F)
obj<-RunPCA(obj)
obj<-IntegrateLayers(obj, method = CCAIntegration, orig.reduction = 'pca', new.reduction='integrated.sct', normalization.method = "SCT", verbose = F)
obj <- FindNeighbors(obj, dims = 1:50, reduction = "integrated.sct")
obj <- FindClusters(obj, resolution = c(1,1.4,1.6,2), cluster.name = c('S1','S2','S3','S4'))
obj<-RunUMAP(obj, dims = 1:50, reduction = "integrated.sct", reduction.name = 'umap.sct')
DimPlot(obj, reduction = "umap.sct", pt.size = 1, label = T)+NoLegend()
obj<-PrepSCTFindMarkers(obj)
Markers<-FindAllMarkers(obj, only.pos = T, min.pct = 0.1)

Thanks!