The field of geometric deep learning has had a profound impact on the
development of innovative and powerful graph neural network architectures.
Disciplines such as computer vision and computational biology have benefited
significantly from such methodological advances, which has led to breakthroughs
in scientific domains such as protein structure prediction and design. In this
work, we introduce GCPNet, a new geometry-complete, SE(3)-equivariant graph
neural network designed for 3D molecular graph representation learning. We
demonstrate the state-of-the-art utility and expressiveness of our method on
six independent datasets designed for three distinct geometric tasks:
protein-ligand binding affinity prediction, protein structure ranking, and
Newtonian many-body systems modeling. Our results suggest that GCPNet is a
powerful, general method for capturing complex geometric and physical
interactions within 3D molecular graphs for downstream prediction tasks. The
source code, data, and instructions to train new models or reproduce our
results are freely available at
this https URL.

介绍了一种新的 GCPNet 图神经网络，其具有 SE (3) 置换对称性，用于 3D 分子图形表示学习，在蛋白质 - 配体结合亲和力预测，蛋白质结构排序和牛顿多体系统建模任务中均取得了最先进的效果。