I have a question about the anndata → Seurat conversion. Currently, the conversion fails if the resulting sparse matrix exceeds ~2^31 non‑zero entries (R’s Matrix package limit). This usually only surfaces when combining the datasets at the end. I know that very large datasets can be handled via BPCells‑backed Seurat objects, but these require absolute paths and are harder to share in a core‑facility setting. Is there a better way to store this size as an Seurat object? (Currently, the process is just turned off, but would be nice to have the conversion included in the future)