We first demonstrate our bridge integration strategy by performing cross-modality mapping on scATAC-seq and scRNA-seq samples of human bone marrow mononuclear cells (BMMCs). These samples consist of cells representing the full spectrum of hematopoietic differentiation, including hematopoietic stem cells, multi and oligopotent progenitors, and fully differentiated cells. As part of HuBMAP, we have leveraged public datasets to construct a comprehensive scRNA-seq reference (‘Azimuth reference’; 297,627 cells) of human BMMC, carefully annotating 10 progenitor and 25 differentiated cell states (Fig. 2a). We aimed to map scATAC-seq ‘query’ datasets of human BMMC43 (16,266 whole bone marrow profiles and 9,893 CD34+ enriched profiles) to this reference (Fig. 2b). We used a 10x Multiome dataset44 (32,368 cells paired snRNA-seq + scATAC-seq) that was publicly released as part of NeurIPS 2021, as a bridge.
