Discovery of Highly Bioactive Peptides through Hierarchical Structural Information and Molecular Dynamics Simulations

Peptide drugs play an essential role in modern therapeutics, but the computational design of these molecules is hindered by several challenges. Traditional methods like molecular docking and molecular dynamics (MD) simulation, as well as recent deep learning approaches, often face limitations related to computational resource demands, complex binding affinity assessments, extensive data requirements, and poor model interpretability. Here, we introduce