DSFAS: Harnessing the Power of Artificial Intelligence and Deep Eutectic Solvents to Boost Antioxidant Synergism

With the support from two ACRE grants (2022 and 2023), our group has been able to develop several artificial intelligence models to predict synergistic interactions of antioxidant. These models considered that these synergistic complexes of antioxidants were supported by hydrogen-bonding, a critical aspect of the formation of deep eutectic solvents. These projects have allowed us to predict and unveil multiple new combinations with synergistic behavior. Moreover, these projects have allowed us to identify and address limitations related to the use of transformer-based neural networks (and SMILES notation) as well as demonstrate the advantages of including experiments that specifically support learning aspects of the chemistry being used. We now turn our attention to the overlap between these two projects and hypothesize that our current models could be applied to form deep eutectic solvents (DES) integrating antioxidants and that such pre-formed complexes would provide enhanced antioxidant activity when deployed in fat samples.

Funding for this project was provided by the USDA-NIFA Program (Garcia, Whitehead)