Researchers from North Carolina State University have demonstrated that molecular dynamics simulations and machine learning techniques could be integrated to create more accurate computer prediction models. These “hyper-predictive” models could be used to quickly predict which new chemical compounds could be promising drug candidates.