Learning from Protein Structure with Geometric Vector Perceptrons
This addresses a gap in machine learning for biomolecular structure analysis, offering a novel approach that could benefit computational biology and drug discovery.
The paper tackled the lack of a unifying network architecture for learning on 3D protein structures by introducing geometric vector perceptrons, which improved performance on model quality assessment and computational protein design over state-of-the-art methods.
Learning on 3D structures of large biomolecules is emerging as a distinct area in machine learning, but there has yet to emerge a unifying network architecture that simultaneously leverages the graph-structured and geometric aspects of the problem domain. To address this gap, we introduce geometric vector perceptrons, which extend standard dense layers to operate on collections of Euclidean vectors. Graph neural networks equipped with such layers are able to perform both geometric and relational reasoning on efficient and natural representations of macromolecular structure. We demonstrate our approach on two important problems in learning from protein structure: model quality assessment and computational protein design. Our approach improves over existing classes of architectures, including state-of-the-art graph-based and voxel-based methods. We release our code at https://github.com/drorlab/gvp.