ran into a curious wrinkle trying to use scDblFinder to find doublets in a sc-RNA-Seq dataset. To do this you first convert your seurat object to a sce object, using as.SingleCellExperiment , which failed. The seurat object is a seurat v5 object with multiple layers, made from merging individual objects. It has to do with seurat v5 using the multiple layers method, (as.SingleCellExperiment works fine on a seurat v5 object with only a single layer.) Searching this issue online showed others having this problem but no working solution that I could find. I figured out a workaround involving extracting and collapsing the individual layers, and adding back to the object as single count and data layers (and removing the multiple layers). It sort of makes a seurat v5 object resemble a v3 object. It's a little bit involved, but not too tricky.
ran into a curious wrinkle trying to use scDblFinder to find doublets in a sc-RNA-Seq dataset. To do this you first convert your seurat object to a sce object, using as.SingleCellExperiment , which failed. The seurat object is a seurat v5 object with multiple layers, made from merging individual objects. It has to do with seurat v5 using the multiple layers method, (as.SingleCellExperiment works fine on a seurat v5 object with only a single layer.) Searching this issue online showed others having this problem but no working solution that I could find. I figured out a workaround involving extracting and collapsing the individual layers, and adding back to the object as single count and data layers (and removing the multiple layers). It sort of makes a seurat v5 object resemble a v3 object. It's a little bit involved, but not too tricky.