tsne not found in reducedDims(data)

[EAC-T]

Dear Kelly,

I have a seurat object, I did tsne and I can plot it using DimPlot function in Seurat. I then did this to run slingshot sce <- as.SingleCellExperiment(seurat_object) sce <- slingshot(sce, clusterLabels = 'new.ident', reducedDim = 'tsne')

I'm getting an error that "tsne not found in reducedDims(data)." I'm sure I have tsne, if I do PCA it works, but I will prefer to do my analysis on tsne.

Any idea why?

Thank you a lot,

[kstreet13]

Hi @EAC-T,

Thanks for the report! This seems to be the same as issue #187. I'm not sure if something changed in how the as.SingleCellExperiment function works, but I haven't noticed a difference. What do you get if you print the Seurat object and the SCE?

Best, Kelly

[EAC-T]

Hi Kelly,

print(seurat) gives me this: An object of class Seurat 15515 features across 1343 samples within 1 assay Active assay: RNA (15515 features, 5000 variable features) 3 dimensional reductions calculated: pca, umap, tsne

print(sce) give me this: class: SingleCellExperiment dim: 15515 1343 metadata(0): assays(2): counts logcounts rownames(15515): Mrpl15 Lypla1 ... Grpr Gm21860 rowData names(0): colnames(1343): WAVEGEXAAACCTGAGATGTTAG-1 WAVEGEXAAACCTGCAGCTATTG-1 ... CTLGEXTTTGTCAAGTAACCCT-1 CTLGEXTTTGTCATCGTGGGAA-1 colData names(13): orig.ident nCount_RNA ... slingshot slingPseudotime_1 reducedDimNames(3): PCA UMAP TSNE mainExpName: RNA altExpNames(0):

Thank you a lot

[kstreet13]

Hi @EAC-T,

Ah, I think this is being caused by the as.SingleCellExperiment() function capitalizing the names of all the dimensionality reductions.

3 dimensional reductions calculated: pca, umap, tsne

reducedDimNames(3): PCA UMAP TSNE

You may just need to run Slingshot like this:

sce <- slingshot(sce, clusterLabels = 'new.ident', reducedDim = 'TSNE')

Let me know if that doesn't solve it! Kelly