Participated · Seed · Apr 2018

ReviveMed aggregates large, fragmented metabolomics and mass-spectrometry data and applies machine-learning models to distinguish metabolites and uncover links between metabolites, proteins, drugs and disease. The company builds a unified database from public and proprietary data, including metabolites that have not been previously analyzed, and has a patent on its process; its foundational work was published in Nature Methods. ReviveMed has demonstrated proof-of-concept results identifying new mechanisms for Huntington’s and therapeutic targets, and its first clinical focus is fatty liver disease. The team conducts hundreds of human tests to improve the model and disambiguate metabolite signals that standard assays miss. Its commercial model is R&D-as-a-service to pharmaceutical companies, with several revenue-generating pilots already in place and dealings with two top-5 pharma firms. The company also won $200,000 in cloud computing credit from a Google machine-learning competition to support its work.