Error in plot_umap

[sanchezy]

Hi all,

I have a SeuratObject sc:

24026 features across 60260 samples within 1 assay 
Active assay: RNA (24026 features, 0 variable features)
2 dimensional reductions calculated: umap, ref.umap

calculate_auc runs fine and I can produce the lollipop plot but when I try to overlay the AUC onto the umap by using: plot_umap(augur, sc, mode = "rank")

I get the following error: Error in$<-.data.frame(tmp, "auc", value = numeric(0)) : replacement has 0 rows, data has 60260

Is there any solution to this? Any help will be greatly appreciated!

[skinnider]

Hi @sanchezy - I'm not entirely sure what's going on but it seems like plot_umap isn't able to match up the AUCs calculated by Augur to the data frame needed to overlay them onto a UMAP. Just a guess, but is it possible that in the Seurat object, your cell types are not in a column named "cell_type"? I'd be happy to take a closer look and see if I can debug whatever's going on here, but I would need you to send the sc object you're using as input to Augur. Are you able to share that?

[sanchezy]

Hi @skinnider, Yes, the cell type annotations are in a column called 'CellTypeManual'. So, I ran the code as:

augur = calculate_auc(sc, label_col = 'Condition2', cell_type_col = 'CellTypeManual')

I made a smaller version of the object sc and will send it by email shortly. Thanks for looking into this.

[skinnider]

Hi @sanchezy, sorry for the delay. I'm not able to reproduce this error with the dataset you sent over when I run the following two commands:

augur = calculate_auc(sc, label_col = 'Condition2', cell_type_col = 'CellTypeManual')
plot_umap(augur, sc, cell_type_col = 'CellTypeManual', mode = 'rank')

Are you passing the cell_type_col argument into plot_umap as well as calculate_auc?

[sanchezy]

Hi @skinnider , thanks a lot for looking into this. Indeed, the solution was to pass the cell_type_col argument into plot_umap. Many thanks!