Dear all,
I am using seurat to analyze spatial data. I have 8 datasets from 2 batches merged together. I aim to use dittoSeq for visualizations, however, I am facing a problem regarding NA values. The batches are saved in the metadata slot of my seurat object. DimPlot from Seurat is working well.
dittoDimPlot(brain.merge, var="Batch")
Error: Cannot find 'NA' in this Seurat object
Best, Nina
Without knowing the package, a quick look at the help page suggested that this function plots a given variable (here Batch) using a dimensionality reduction.
The relevant argument default is reduction.use = .default_reduction(object). If the object has none, then it is NA. Hence, I assume your object has not been run through any reductuon before.
What is the output of: dittoSeq:::.default_reduction(brain.merge)
Seurat is not a Bioconductor package, so this question is off-topic for this site. You should ask your question on the Seurat Discussions page.
traceback()
8: stop("Cannot find '", i, "' in this Seurat object", call. = FALSE)
7: [[.Seurat(object, reduction)
6: object[[reduction]]
5: Embeddings(object = object[[reduction]], ...)
4: Embeddings.Seurat(object, reduction = reduction.use)
3: Seurat::Embeddings(object, reduction = reduction.use)
2: .extract_Reduced_Dim(reduction.use, dim.1, object)
1: dittoDimPlot(brain.merge, var = "Batch")
I never used traceback before, but this was a good hint to search for a solution. In this data set there was no UMAP projection calculated before. After RunUMAP it was working =) thanks for your help!
Best, Nina
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
You're right. After RunUMAP it changes to "umap" and it is working without any error.
Thanks!
Great, you're welcome. I submitted a PR to dittoSeq GitHub with a fix for this: https://github.com/dtm2451/dittoSeq/pull/123
I've updated the fix and just pushed through to both the release and devel versions of the package! It'll take a few days to go through.
Ultimately, running RunUmap and then re-running dittoDimPlot() was indeed your correct path... ATpoint and my fix just makes it so you'd actually have gotten the error message -- "No dimensionality reduction slots in 'object'. Add one, or provide embeddings directly to 'reduction.use'." -- which was intended to point you in that direction.