PanKgraph is the PanKbase Knowledge Graph, a biomedical KG focused on the human pancreas and diabetes research. It represents PanKbase’s goal of integrating diverse type 1 diabetes datasets with expert-curated knowledge in a centralized, open-source and computation-ready resource. The attached entity model shows a Biolink- and ontology-aligned graph with 14 classes and 117 predicates, including genes, diseases, cells, cell lines, articles, Gene Ontology terms, SNPs, SNVs, open chromatin regions, datasets and RDF statements. Its relationships and edge attributes support biological interpretation across gene expression, regulatory genomics, genetic variation, fine-mapping and QTL evidence, gene-condition associations, gene-gene interactions, publication provenance and dataset metadata. PanKgraph is designed to connect previously disjointed entities such as genes, SNPs and pancreatic eQTLs, helping researchers explore mechanisms of diabetes pathogenesis. As part of the broader PanKbase/HIRN ecosystem, it supports FAIR, reproducible, collaborative analysis of pancreas multi-omics, donor-derived data and disease-relevant evidence for biomarker and therapeutic target discovery.
