Error with ImageDimPlot

[Chengwei94]

Hi, I am trying out the new image spatial toolkit with my own dataset, and I met with some problems

wt1.crop <- Crop(seurat_ns[["fov1"]], x = c(-6000, 0), y = c(-25000, -30000))
seurat_ns[["fov_wt1"]] <- wt1.crop
DefaultBoundary(seurat_ns[["fov_wt1"]]) <- "centroids"
DefaultAssay(seurat_ns) <- "Nanostring"
ImageDimPlot(seurat_ns, fov = "fov_wt1", group.by = "seurat_clusters")

ImageDimPlot(seurat_ns, fov = "fov_wt1", group.by = "seurat_clusters")
Error in df[df$cell %in% cells, , drop = FALSE] :
incorrect number of dimensions

I have checked the Cells(seurat_ns) and Cells(fov_wt1) and they overlap
Addon: It seem that the error is with the underscore in the assay name, since theres a str_split by underscore in the code (Not sure if its intended)

Furthermore , I found that the LoadNanostring needs the segmentation file to work. Is there a way to obtain the segmentation file from the images?

[AustinHartman]

Hi, thanks for raising this. Please see the latest commit of the feat/imaging branch to find a fix supporting underscores in fov names. Let me know if that solves your issue.

Currently there is not a way to obtain segmentations directly from images, but the nanostring segmentations are available for download here