Integrating D-MPNN and molecular dynamics for chemical interaction prediction

A novel hybrid approach combining Directed Message Passing Neural Networks and molecular dynamics simulations to predict chemical interactions with substrates for designing innovative cosmetic and detergent formulations.