What problem does it solve? Annotating an scATAC-seq query is hard when no matched RNA data exists for it, but a well-curated scRNA reference does. This Skill provides two Seurat workflows — Azimuth reference mapping and bridge integration — to transfer cell-type labels from an RNA reference onto a query, including ATAC-only queries connected through a paired multiome bridge dataset. ## Core Features & Use Cases - Azimuth Mapping: Map RNA-only (or WNN-integrated) queries onto SeuratData references such as pbmcref with RunAzimuth, yielding predicted cell-type labels and mapping scores. - Bridge Integration: Use FindBridgeTransferAnchors and MapQuery to project an ATAC-only query through a multiome bridge dataset into an RNA reference, transferring l1/l2 labels and reference UMAP coordinates. - Quality Control Guidance: Apply conventional score thresholds (predicted score and mapping score above 0.5) and troubleshoot pitfalls like mismatched peak sets, low overlap, and memory blow-ups. - Use Case: You have an scATAC PBMC dataset with no RNA modality and want cell-type annotations. Load a PBMC multiome bridge and the pbmcref RNA reference, find bridge transfer anchors, and map predicted.l2 labels onto your ATAC cells. ## Quick Start Ask the AI to map cell-type labels from the pbmcref RNA reference onto your ATAC-only Seurat object using bridge integration with a PBMC multiome dataset.
