Jianan Li

Protocol for Membrane Permeability Prediction of Cyclic Peptides using Descriptors Obtained from Extended Ensemble Molecular Dynamics Simulations and Chemical Structures

The paper develops a machine learning protocol to predict cyclic peptide membrane permeability using 3D descriptors from MD simulations and 2D chemical descriptors, achieving R = 0.77 and RMSE = 0.62 on training data.

CycPeptMP: Enhancing Membrane Permeability Prediction of Cyclic Peptides with Multi-Level Molecular Features and Data Augmentation

CycPeptMP predicts cyclic peptide membrane permeability using multi-level features and data augmentation, achieving high accuracy with RMSE 0.503 and correlation 0.883.