Introduction to the Atlas

Introduction to the KPMP Kidney Tissue Atlas

The goal of the Kidney Precision Medicine Project (KPMP) is to understand kidney disease in mechanistic terms in order to find new ways to treat them. KPMP researchers use kidney biopsies to build multiple maps of the kidney, together called the Kidney Tissue Atlas. These maps show the important cells, regions, and disease pathways that lead to CKD and AKI. The Atlas is a free set of interactive tools that helps the research community retrieve, explore, and analyze KPMP data. Its data comes from KPMP participant biopsies and reference tissue samples, in both raw and processed form.

‍

The Dashboard (home page)

The home page is the entry point to the Atlas's three tools as well as links to other tools and resources from our collaborators:

• Explorer searches markers or cell types and shows summary visualizations across KPMP's omics technologies.
• Repository lets you download data generated from participant biopsies and reference tissue.
• Spatial Viewer displays interactive spatial data from KPMP's imaging technologies.

The Atlas Data Summary shows how many participants contributed to each omics type. The counts are broken down by enrollment category: healthy reference, CKD, AKI, diabetes mellitus–resilient (DM-R), and all participants. The summary also explains that a subset of the Repository's raw data has been analyzed and loaded into Explorer and Spatial Viewer for interactive use. Check this table first to see what data exists before choosing a tool.

‍

Explorer

Explorer is a search engine for markers and cell types. It shows summary visualizations for single-nucleus RNA-seq, single-cell RNA-seq, regional transcriptomics, and regional proteomics data.

You can search by gene or by cell type:

• Gene search, single-cell or single-nucleus data. You get three views: a reference UMAP showing how the full dataset was clustered and the inferred cell type of each cluster, a feature plot of the gene's expression, and a table comparing the gene's expression in each cluster against all other clusters. You can filter to AKI, CKD, or healthy reference samples.
• Gene search, regional transcriptomics data. You get a bubble plot of the gene's expression across tissue types and microdissected regions, plus a table comparing each region against all others.
• Cell type search. You can type a cell type or pick one from a diagram. Explorer then shows all data mapped to that cell type and lists the most differentially expressed genes for the cluster or region you select. The diagrams use detailed illustrations from the Human Reference Atlas.

Explorer also links to CELLxGENE for disease-specific differential expression analysis.

‍

Repository

The Repository provides access to all dataset files KPMP generates, and KPMP adds new files as they pass quality control. Files are either open access or controlled access:

• Open access files download directly.
• Controlled access files require a data use agreement between KPMP and your institution.

Filters are grouped into participant metadata, clinical metadata, and dataset metadata. A Recently Released option shows only the latest data packages. Multi-select filters let you narrow results to a specific cohort and data type. You can also filter by DOI to find the exact data behind a KPMP publication. When your list is ready, you can download it as a CSV for data requests or internal tracking.

‍

Spatial Viewer

Spatial Viewer lets you search KPMP's spatial datasets by data-level and participant-level attributes and then view them.

Finding a dataset. The filter panel has two tabs, DATASET and PARTICIPANT. Select a sample ID to open that dataset. Select a participant ID to see that participant's clinical information and the other Atlas data available for them.

Viewing a dataset. Spatial Viewer uses the Vitessce visualization platform, and the layout adapts to each data type:

• Whole slide images. Zoom and download.
• Multichannel images such as CODEX. Choose markers and channel colors.
• Spatial transcriptomics. Search a gene to see expression as a tissue heatmap linked to a UMAP scatterplot. Lasso a group of points in either view to highlight the same spots in the other.
• Spatial metabolomics, lipidomics, and N-glycomics. These open in the external METASPACE platform.

‍