Single-cell Analysis Tool Inputs
Cellismo / Muon file: [sample_name].cellismo
Seurat file: [sample_name]_seurat.rds
These are prebuilt input files to popular single-cell analysis toolkits:
- Cellismo / Muon (bdbiosciences.com/cellismo) / (muon.scverse.org)
and
- Seurat v4 (satijalab.org/seurat/)
In RNA and AbSeq experiments, each file contains the RSEC molecules-per-cell data table for putative cells, along with cell and bioproduct metadata.
In ATAC-Seq experiments, each file contains the cell-by-peak data table for putative cells, and a cell-by-gene_activity table calculated based on peak region annotations, along with cell and bioproduct metadata. If Transcription Factor Motif analysis was performed, a cell-by-motif table of motif enrichment z-scores generated by the pyChromvar algorithm is also included, and a peak-by-motif matrix indicating which motifs were detected in which peaks.
Metadata includes (if applicable): sample tag calls, putative cell origin, TCR/BCR chain types and CDR3, Immune cell type (Experimental), protein aggregate, and tSNE/UMAP coordinates.
Additionally, the chromosome and contig lengths of the reference genome used are attached to the .cellismo object (mdata.uns['genome_contig_lengths']).
The names of the chromosomes/contigs have unusual characters encoded to be "URL-safe" using the function urllib.parse.quote_plus() from the Python package urllib.parse.
These are automatically decoded when used by the Cellismo viewer, but should be noted if another tool is used to analyze the file.
| Data Type | Seurat Object Location | Cellismo Object (MuData) Location |
|---|---|---|
| Expression/Count Matrices | ||
| mRNA Expression (RSEC) | obj@assays[['RNA']]$counts | mdata.mod['rna'].X |
| AbSeq Expression (RSEC) | obj@assays[['ADT']]$counts | mdata.mod['prot'].X |
| ATAC-Seq Peaks | obj@assays[['peaks']]$counts | mdata.mod['atac_peaks'].X |
| ATAC Gene Activity Scores | obj@assays[['gene_activity']]$counts | mdata.mod['atac_gene_activity'].X |
| ATAC TF Motif Scores (ChromVAR) | obj@assays[['chromvar']]$data | mdata.mod['atac_motif'].X |
| Cell-Level Metadata | ||
| Sample Tag Assignments, VDJ Per-Cell Info, Immune Cell Type Predictions, Protein Aggregate Status | obj@meta.data A data frame where columns correspond to metadata fields (e.g., obj$Sample_Name, obj$TCR_Paired_Chains). | mdata.mod[MODALITY].obs Data linked to each modality. A data frame where columns correspond to metadata fields (e.g., mdata.mod['rna'].obs['Cell_Type_Experimental']). |
| Dimensionality Reduction (UMAP/tSNE) | obj@reductions$umap or obj@reductions$tsne | mdata.mod[MODALITY].obsm['X_umap'] or mdata.mod[MODALITY].obsm['X_tsne'] Data linked to each modality. |
| Feature-Level Metadata | ||
| ATAC Peak-to-Gene Links | Not present | mdata.mod['atac_peaks'].uns['atac']['peak_annotation'] |
| ATAC TF Motif Matrix (Peak x Motif) | Stored in a Motif object associated with the peaks assay: obj@assays[['peaks']]@motifs | mdata.mod['atac_peaks'].varm['peaks_by_motif'] Motif names are stored in mdata.mod['atac'].uns['atac']['motif_names'] |
| ATAC TF Motif Genomic Positions | Stored in a Motif object associated with the peaks assay: obj@assays[['peaks']]@motifs@positions | mdata.mod['atac_motif'].uns['motif_positions'] |