Supporting information for paper titled "Computing and visualizing cells of 1,183 types in 1,558 anatomical structures of the Human Reference Atlas" (WORK IN PROGRESS)

The Human Reference Atlas (HRA) aims to capture the number and type of different cell type populations in the 1000s of 3D anatomical structures that make up the human body. Using a combination of experimental data from single-cell studies and ASCT+B tables that capture the relationships between anatomical structures (AS), cell types (CT), and biomarkers (B), we present a general workflow that can be applied to compute and visualize 1,000,000,000,000 cells of 1,183 types in 1,558 anatomical structures defined by the Human Reference Atlas v1.3. While the initial layout of the cells is random, it is valuable for (1) supporting spatial search (e.g., the retrieval of “3D corridors” that contain certain sets of cell types), (2) registering tissue blocks for which spatial metadata is missing into the 3D HRA, (3) visual identification and exploration of anatomical structures that are homogenous or heterogenous, and (4) explaining HRA construction and usage to a general audience via the HRA Organ Gallery in VR. The paper concludes with a validation study that compares HRA cell type populations with prior experimental data and existing BioNumbers.