Molecular distance matrix prediction based on graph convolutional networks

Molecular structure has important applications in many fields. For example, some studies show that molecular spatial information can be used to achieve better prediction results when predicting properties. However, traditional geometry calculations, such as density functional theory (DFT), are time-consuming. In view of this, we propose a model based on graph convolutional networks predict the pairwise distance between atoms, also called matrix molecule (DMGCN). order indicate effect DMGCN model, is compared with DeeperGCN-DAGNN and method calculating conformation RDKit. Results MAE smaller than addition, distances predicted by calculated QM9 dataset properties, thus showing effectiveness model. Our code available at https://github.com/Lin12138xh/DMGCN.