Moliere: Automatic Biomedical Hypothesis Generation
Justin Sybrandt, Michael S. Shtutman, Ilya Safro · TigerPrints (Clemson University) · 2019
Medical research is expensive and risky. Drug manufacturers need to prioritize their early investments with very little experimental data in order to most efficiently discover new treatments. Moliere is a hypothesis generation system that identifies implicit yet-unknown connections already present within the body of medical literature. Using our discovery and ranking system, medical researchers can identify fruitful research directions earlier in the discovery process, before time consuming and expensive experiments. For example, we used Moliere to identify a new gene-treatment target for HIV-associated Neurodegenerative Disease, which we later confirmed in laboratory experiments.