Error when running convert2anndata even though I have all 3 packages installed

[alifarhat40]

Hello. I am running into the following error when running the function to convert in the command line. And when I try to do it in R it takes forever to run. In the terminal, I get the following:

`Rscript -e "convert2anndata::cli_convert()" -i scATACSeurat.rds -a ATAC -o scATACSeurat.h5ad
[2025-02-07 21:43:36] Loading data from: scATACSeurat.rds
[2025-02-07 21:44:06] Input object class: Seurat
[2025-02-07 21:44:06] Converting Seurat object...
[2025-02-07 21:44:06] Proceeding with custom conversion...
[2025-02-07 21:44:06] Found assays: ATAC
[2025-02-07 21:44:06] Processing assay: ATAC
[2025-02-07 21:44:06] Found 2 layers: counts, data
[2025-02-07 21:44:18] Attempting to preserve dimensional reductions.
[2025-02-07 21:44:18] Updated reduction ' lsi ' to match combined cells.
[2025-02-07 21:44:19] Updated reduction ' umap ' to match combined cells.
[2025-02-07 21:44:21] Found 2 graphs. Moving to colPair.
[2025-02-07 21:44:24] Custom conversion to SingleCellExperiment completed successfully.
[2025-02-07 21:44:24] Data loaded and converted successfully if needed.
[2025-02-07 21:44:24] Summary of SingleCellExperiment object:

class: SingleCellExperiment
dim: 298815 48229
metadata(2): version commands
assays(2): ATAC_counts ATAC_data
rownames(298815): chr1:3119609-3120608 chr1:3120985-3121984 ...
chrX:170675661-170676660 chrX:170676888-170677887
rowData names(0):
colnames(48229): Multiome-Old-HSC-rep1-AAACAGCCAACTAGAA-1
Multiome-Old-HSC-rep1-AAACAGCCAAGGTGCA-1 ...
Multiome-Young-LinNeg-rep1-TTTGTTGGTCCCGAAG-1
Multiome-Young-LinNeg-rep1-TTTGTTGGTGCTTTAC-1
colData names(12): orig.ident nCount_ATAC ... seurat_clusters cellType
reducedDimNames(2): lsi umap
mainExpName: NULL
altExpNames(0):

[2025-02-07 21:44:26] Checking for altExperiments...
[2025-02-07 21:44:26] Processing assay 'ATAC' for anndata.X...
[2025-02-07 21:44:26] WARNING: The specified assay 'ATAC' is not available. Using 'ATAC_counts' as active layer instead.
[2025-02-07 21:44:45] Using 'ATAC_counts' assay as the main data matrix.
[2025-02-07 21:44:45] Processing all assays...
[2025-02-07 21:45:24] Assays processed.
[2025-02-07 21:45:24] Processing dimensional reductions...
[2025-02-07 21:45:26] Dimensional reductions processed.
[2025-02-07 21:45:26] WARNING: Failed to extract obs/colData with internal=TRUE.
[2025-02-07 21:45:26] Successfully extracted obs/colData without internal=TRUE.
[2025-02-07 21:45:26] WARNING: The following columns are missing without internal=TRUE and some meta information might be lost: reducedDims, altExps, colPairs
[2025-02-07 21:45:26] Successfully extracted var/rowData with internal=TRUE.
Error in mcols(pairs[[name]]) : could not find function "mcols"
Calls: -> convert_to_anndata -> extract_pairs
Execution halted`

Any idea what this could be? I have all the packages installed.

[katosh]

Hi @alifarhat40,

Thank you for bringing this up! I can see how this issue could arise. In commit ca4fc6f we now explicitly ensure that the mcols function is properly loaded.

That said, I wasn’t able to reproduce the error on my end. Could you please test with the latest changes and let me know if this resolves your issue? If the problem persists, any additional details about your setup would be helpful.

Looking forward to your feedback!