DeepMet usage by Michael Skinnider

Dr. Michael Skinnider, Assistant Professor at Princeton University, presented DeepMet, a chemical language model trained on the structures of all known human metabolites. The model expands the chemical space of known human metabolites with new, metabolite-like structures and predicts the existence of as-of-yet undiscovered metabolites.

It was demonstrated how DeepMet can be applied to predict metabolites, which can be synthesized and then lead to their targeted discovery. A direction of interest for HyperMet is that significant changes in metabolomics between hypertrophic/atrophic/normal muscle appear, and unidentified metabolites could be determined.

HyperMet research examines the impact of muscle growth (hypertrophy) and muscle loss (atrophy) on metabolism. Increased muscle mass reduces the risk of obesity, diabetes, osteoporosis, and potentially cancer. We are exploring the underlying metabolic processes to develop new strategies for prevention and everyday life.