Abstract
Knowledge graphs (KG) are an emerging approach to organize biomedical research data by abstracting knowledge into networks that connect genes, diseases, pathogens, drugs, metabolites, cell types, pathways, patients, researchers, and other related concepts. While there are many KG databases, most do not offer a free and customizable web-based user interface (UI) to query and interact with KG. We developed an open-source UI to create interactive websites from data stored in KG databases. So far, the KG-UI has been applied to eight bioinformatics projects: ReproTox-KG, Enrichr-KG, Harmonizome-KG, Data Distillery KG-UI, Biomarker-KG, Common Fund Data Ecosystem Gene Set Enrichment (CFDE-GSE), ChEA-KG, and lncRNAlyzr. To demonstrate how to install and customize the KG-UI, we created a demo KG that displays a network of relationships between NIH-funded principal investigators (PIs) from a single institution in a specific year. This PI network was constructed by querying PubMed using grant information collected by the Blue Ridge Institute for Medical Research (BRIMR). After publications were collected from PubMed, PI name disambiguation and metadata formatting was performed to create the PI demo KG. Instructions for ingesting this network of PIs into a KG database, setting up the web app, and customizing the app are provided in a step-by-step user guide. The KG-UI source code is available from: https://github.com/MaayanLab/Knowledge-Graph-UI/ and the demo KG is available from: https://github.com/MaayanLab/PINetworkDemo. © 2025 The Author(s). Current Protocols published by Wiley Periodicals LLC. Basic Protocol 1: Creating the principal investigator network Basic Protocol 2: Validating the principal investigator network Basic Protocol 3: Ingesting the KG assertions into Neo4j Basic Protocol 4: Interacting with the KG via the Neo4j remote interface Basic Protocol 5: Building and customizing the KG-UI Basic Protocol 6: Interacting with the KG-UI website.