Neural Network-Derived Potts Models for Structure-Based Protein Design using Backbone Atomic Coordinates and Tertiary Motifs
A. J. Li, M. Lu, I. T. Desta, V. Sundar, G. Grigoryan, A. E. Keating
This paper combines neural network models with tertiary motif data to improve protein design by generating Potts energy models that predict folding and energetics.
bioRxiv · Aug 2, 2022Biophysics