Geometric Vector Perceptrons, or GVPs, are a type of neural network layer commonly combined with Graph neural networks for use on Inverse folding problems. Jing et al. (1) used them for computational protein design and model quality assessment. They were also used by ESM-IF (2).
Figures
Figure 1A from (1)
Notes
(3) found that GVP-GNN achieved greater sequence recovery than standard GNNs of the same size with the same test set, but did not show improvements in downstream prediction.
1.
Jing B, Eismann S, Suriana P, Townshend RJL, Dror R. Learning from Protein Structure with Geometric Vector Perceptrons. In: ICLR 2021. 2021. Available from: https://openreview.net/forum?id=1YLJDvSx6J4
2.
Hsu C, Verkuil R, Liu J, Lin Z, Hie B, Sercu T, et al. Learning inverse folding from millions of predicted structures. openRxiv; 2022. Available from: https://doi.org/10.1101/2022.04.10.487779
3.
Yang KK, Zanichelli N, Yeh H. Masked inverse folding with sequence transfer for protein representation learning. Protein Engineering, Design and Selection. 2022;36. Available from: https://doi.org/10.1093/protein/gzad015