Talk2Biomodels and Talk2KnowledgeGraph: AI agent-based application for prediction of patient biomarkers and reasoning over biomedical knowledge graphs

Gurdeep Singh, Lilija Wehling, Ahmad Wisnu Mulyadi, Rakesh Hadne Sreenath, Thomas Klabunde, Tommaso Andreani, Douglas McCloskey

Track: Tiny paper track (up to 4 pages)

Abstract: In this study, we present Talk2Biomodels (T2B) and Talk2KnowledgeGraphs (T2KG) as open-source, user-friendly, large language model-based agentic AI platforms designed to democratize access to computational models of disease processes using natural language. T2B and T2KG eschew the traditional graphical user interface (GUI) and minimally adaptable workflow in favour of a modern agentic framework to provide a dynamic and immersive experience to explore the biology of disease in silico and how different treatment options can be efficacious in different virtual patient populations. T2B supports models encoded in the open-source community format Systems Biology Markup Language (SBML) for quantitative prediction of patient biomarkers and integrates with biomedical knowledge graphs to provide qualitative insights not captured in the computational model. A use case in precision medicine is presented to demonstrate how experts and non-experts in computational biology and data science can benefit from T2B and T2KG.

Submission Number: 95