We consider the prediction of interfaces between proteins, a challenging problem
with important applications in drug discovery and design, and examine the
performance of existing and newly proposed spatial graph convolution operators
for this task. By performing convolution over a local neighborhood of a node of
interest, we are able to stack multiple layers of convolution and learn effective
latent representations that integrate information across the graph that represent the
three dimensional structure of a protein of interest. An architecture that combines
the learned features across pairs of proteins is then used to classify pairs of amino
acid residues as part of an interface or not. In our experiments, several graph
convolution operators yielded accuracy that is better than the state-of-the-art SVM
method in this task.
evancofer
commented
6 years ago
We consider the prediction of interfaces between proteins, a challenging problem with important applications in drug discovery and design, and examine the performance of existing and newly proposed spatial graph convolution operators for this task. By performing convolution over a local neighborhood of a node of interest, we are able to stack multiple layers of convolution and learn effective latent representations that integrate information across the graph that represent the three dimensional structure of a protein of interest. An architecture that combines the learned features across pairs of proteins is then used to classify pairs of amino acid residues as part of an interface or not. In our experiments, several graph convolution operators yielded accuracy that is better than the state-of-the-art SVM method in this task.
https://papers.nips.cc/paper/7231-protein-interface-prediction-using-graph-convolutional-networks.pdf