search

cole-trapnell-lab / monocle3

Other
318 stars 98 forks source link

Cluster_cells : Error in graph.adjacency.dense #718

Open Dooo0k opened 2 months ago

Dooo0k commented 2 months ago

My seurat object has 244 cells, and when I proceed to cluster_cells after converting to cds, I get the following error

cds <- cluster_cells(cds) Error in graph.adjacency.dense(adjmatrix, mode = mode, weighted = weighted, : At vendor/cigraph/src/constructors/adjacency.c:535 : Adjacency matrix should be symmetric to produce an undirected graph. Invalid value

Please let me know how to solve it.

AndGob commented 1 month ago

Hi @Dooo0k, I encountered the same error. Did you deal with that? Thanks

ChrisSteel-bio commented 1 month ago

I'm getting the same error. Anyone figure out a solution? Cheers

AndGob commented 1 month ago

Hi @ChrisSteel-bio, have you solved? If it may help, I had not figure out the problem but remove assays other than 'count' allow me to proceed over. I was working with a cds created using SeuratWrappers::as.cell_data_set(). After that I removed the logcounts assay and done all the analysis in monocle.

rachadele commented 2 weeks ago

experienced the same error running find_gene_modules; what fixed it for me was specifying PCA as the preprocess_method. maybe specifying your desired reduction_method input to the function call would help?

JFanbio commented 1 week ago

https://groups.google.com/g/monocle-3-users/c/tUA7FTeukQY Setting a smaller k will help to solve it.