H-Packer: Holographic Rotationally Equivariant Convolutional Neural Network for Protein Side-Chain Packing

TL;DR

H-Packer is a novel, efficient neural network model that predicts protein side-chain conformations by leveraging rotational symmetry, outperforming traditional methods and competing well with other deep learning approaches.

Contribution

This work introduces H-Packer, a two-stage rotationally equivariant neural network for protein side-chain packing, incorporating symmetry considerations and demonstrating competitive performance.

Findings

H-Packer is computationally efficient.

H-Packer outperforms conventional physics-based algorithms.

H-Packer is competitive with other deep learning methods.

Abstract

Accurately modeling protein 3D structure is essential for the design of functional proteins. An important sub-task of structure modeling is protein side-chain packing: predicting the conformation of side-chains (rotamers) given the protein's backbone structure and amino-acid sequence. Conventional approaches for this task rely on expensive sampling procedures over hand-crafted energy functions and rotamer libraries. Recently, several deep learning methods have been developed to tackle the problem in a data-driven way, albeit with vastly different formulations (from image-to-image translation to directly predicting atomic coordinates). Here, we frame the problem as a joint regression over the side-chains' true degrees of freedom: the dihedral $\chi$ angles. We carefully study possible objective functions for this task, while accounting for the underlying symmetries of the task. We propose Holographic Packer (H-Packer), a novel two-stage algorithm for side-chain packing built on top of two light-weight rotationally equivariant neural networks. We evaluate our method on CASP13 and CASP14 targets. H-Packer is computationally efficient and shows favorable performance against conventional physics-based algorithms and is competitive against alternative deep learning solutions.