Anndata vs cloupe

Question

[Rika]

Hi! I have anndata object of scrna-seq, which was converted to seurat then to cloupe to visualize with loupe browser 8. When converting to seurat, I kept log normalized data since anndata allows users to keep multiple layers of the data, but only one layer for seurat. When converted to cloupe and visualize in loupe, I realized that cell counts expressing gene x were different. I could not figure out why - been stuck on this for hours. Does anyone have any idea why? e.g. there were 6773 cells expressing Ebf2 when using anndata and scanpy, but only 4288 when using loupe.