Geometric Deep Learning for Protein-Protein Interaction Predictions

This work introduces novel approaches, based on geometrical deep learning, for predicting protein-protein interactions. A dataset containing both interacting and non-interacting proteins is selected from the Negatome Database. Interactions are predicted from a graph representing the proteins' t...

Full description

Saved in:
Bibliographic Details
Published in:IEEE access 2022, Vol.10, p.90045-90055
Main Authors: Lemieux, Gabriel St-Pierre, Paquet, Eric, Viktor, Herna L, Michalowski, Wojtek
Format: Article
Language:English
Subjects:
Artificial neural networks
Computer architecture
Convolutional neural networks
Deep learning
geometric deep learning
Graph convolutional networks
graph convolutional neural network
Graphical representations
heat kernel
Heating systems
macromolecular surface
Protein > protein interaction
Proteins
spectral convolutional neural network
Surface waves
wave kernel
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This work introduces novel approaches, based on geometrical deep learning, for predicting protein-protein interactions. A dataset containing both interacting and non-interacting proteins is selected from the Negatome Database. Interactions are predicted from a graph representing the proteins' three-dimensional macromolecular surfaces. The nodes are described with heat and wave kernel signatures. Twenty-one neural network architectures are proposed and compared; these are based on graph convolutional neural networks, spectral convolutional neural networks, and a novel spatio-spectral spatialized-gated convolutional neural network. The experimental results demonstrate the accuracy and the efficiency of the proposed architectures.
ISSN:2169-3536
2169-3536
DOI:10.1109/ACCESS.2022.3201543