Single-Cell Multiome-Rhapsody Service

Figure 1. Integrated RNA and ATAC quality control metrics in mouse brain Multiome-Rhapsody data.
Mouse brain nuclei (n = 4) are projected onto a weighted nearest neighbor (WNN) UMAP generated from the integration of RNA expression and chromatin accessibility profiles. The upper panel displays RNA quality control metrics, while the lower panel shows ATAC quality metrics. RNA QC includes total RNA counts per cell, indicating consistent sequencing depth; number of detected genes per cell, reflecting transcriptome complexity; and the percentage of mitochondrial transcripts, serving as an indicator of cellular stress or damage across the dataset. ATAC QC metrics include total ATAC fragments per cell, reflecting uniform depth of chromatin accessibility; the number of accessible peaks detected per cell, indicating chromatin accessibility complexity; and nucleosome signal, which measures fragment periodicity and is used to assess chromatin organization. Transcription start site (TSS) enrichment scores are also shown, where higher values indicate strong enrichment of signal at promoter regions, consistent with high-quality ATAC data. This figure demonstrates high-quality, consistent RNA and ATAC measurements across the samples.

Figure 2. Integration of RNA and ATAC modalities reveals consistent clustering and reproducible sample composition in mouse brain samples.
Mouse brain nuclei (n = 4) were embedded using a weighted nearest neighbor (WNN) UMAP generated from the joint integration of RNA expression and chromatin accessibility profiles. Left: cells are colored by sample origin, showing an even distribution of each sample across clusters, indicating consistency across samples. Middle: unsupervised clustering identifies 20 transcriptionally and chromatin-accessibility–defined cell populations (clusters 0–19). Right: stacked bar plots display the fraction of cells from each cluster across samples, demonstrating broadly consistent cell-type composition among replicates.

Figure 3. Cell clustering is reproducible across mouse brain samples in Single-Cell Multiome-Rhapsody data
Mouse Brain nuclei (n=4) are visualized using a weighted nearest neighbor (WNN) UMAP embedding derived from integrated RNA expression and chromatin accessibility profiles. Each panel shows cells from a single sample (Rep1, Rep2, Rep3 Rep4), colored by the global cluster assignments (clusters 0–19) obtained from the integrated analysis. The consistent spatial distribution of clusters across all samples indicates reproducible identification of cell populations and minimal sample-to-sample variability between replicates. This pattern further supports the robustness of the joint RNA–ATAC integration and clustering results.

Figure 4. Cluster-resolved chromatin accessibility profiles at representative gene loci in the mouse brain Single-Cell Multiome-Rhapsody dataset.
Aggregated ATAC-seq signal tracks are shown across cell clusters (0–19) at two representative loci, S1pr1 (left, chr3) and Mag (right, chr7). For each cluster, chromatin accessibility signal is normalized and plotted across the genomic region, highlighting cluster-specific accessibility patterns. Gene annotations are shown below each locus, with gene bodies and transcriptional orientation indicated, and called ATAC peaks displayed as gray boxes. Distinct accessibility patterns across clusters suggest cell-type–specific regulatory activity at these loci, with increased accessibility near the S1pr1 and Mag gene regions in specific clusters, consistent with differential regulatory programs across brain cell populations. Violin plots on each panel indicate which clusters express the gene of interest, while the ATAC tracks show where in the genome that accessibility is occurring.