Hi,

Just wondering if there is a convenience function somewhere in the SCE/SFE/SPE/SE object that allows for pulling out feilds from metadata + expression + reduced dimension coordinates all in one command? ie. For when I want to plot a custom umap with gene expression split by sample group, but with more control than the plot_UMAP function typically allows.

Essentially the direct bioconductor equivalent of Seurat's FetchData() (I wasted ages doing that with custom functions before I realised that was inbuilt!)

I know how to build such a table myself, but since its a common pattern I'm wondering if there's a shortcut?

I imagine it would look like this?

get_table_of_data(sce,
     from_metadata=c('timepoint','individual','cluster'),
    from_reddim=list(UMAP='UMAP_1', 'UMAP_2'), 
    from_assay=list(counts=c('CD63','Hba1'))
)

I couldn't see anything in the reference docs, but I'm not sure I'm looking in the right place! Thanks.

This question was mentioned in the #tidiness_in_bioc Bioconductor slack channel. I try to give a tidyomic solution

library(tidySingleCellExperiment)

sce |>
   join_features(c('CD63','Hba1'))

This can be directly piped to ggplot or directly converted to tibble

... |> ggplot()
... |> as_tibble()

Hi Sarah,

scuttle::makePerCellDF() / scuttle::makePerFeatureDF() and scater::ggcells() / scater::ggfeatures(), which build upon the former, will give you what I think you want for SE and SCE. scuttle/scater were designed for single-cell rather than spatial data, but quick testing shows that they may be able to give you the non-spatial components of an _SPE_ and _SFE_. Perhaps you can use these as a starting point or some of the SPE and SPF authors can chime in: dario.righelli, helucro, davide risso, Lukas Weber, lambda

Cheers, Pete