Seurat: Restore support for diffusion maps

[RoganGrant]

I'm wondering if you intend to begin supporting Diffusion Map plotting again with Seurat in the near future. Writing custom functions for this purpose is not particularly difficult, but they are likely to break as Seurat is updated.

[mojaveazure]

Hi Rogan,

Diffusion map visualization can be done with DimPlot, assuming store it as a DimReduc object in the Seurat object, by passing the name of the diffusion map reduction to the reduction parameter.

We currently don't have a timeline for restoring RunDiffusion, but that should be easy for a stable custom function using Seurat's getters and setters. Please see our Developer's Guide for more details about the objects in Seurat v3 and how to interact with them.

[laijen000]

Hello!
I would like to store my diffusion map results as a custom dimensional reduction in Seurat. I looked at the tutorial's example for MDS, but am having a hard time getting it to work with my destiny results.

tmp <- data.matrix(data.frame(DC1 = eigenvectors(dm)[, 1],
DC2 = eigenvectors(dm)[, 2]))
micro[["dm"]] <- CreateDimReducObject(embeddings = tmp, key="DC_", assay=DefaultAssay(micro))
DimPlot(micro, reduction="dm")

When I try to plot, I get the error: Error in Embeddings(object = object[[reduction]])[cells, dims] :
subscript out of bounds

Has anyone been able to store destiny results as a custom dimensional reduction and have advice on how to do so? Thank you so much!