Deeply learning molecular structure-property relationships using attention- and gate-augmented graph convolutional network
This work addresses molecular engineering for materials and drug discovery by enhancing deep learning methods, though it is incremental as it builds on existing GCNs.
The researchers tackled the problem of predicting molecular properties by improving graph convolutional networks (GCNs) with attention and gate mechanisms, resulting in better feature extraction and accurate predictions for properties like solubility and photovoltaic efficiency, with the model identifying key molecular parts for high efficiency.
Molecular structure-property relationships are key to molecular engineering for materials and drug discovery. The rise of deep learning offers a new viable solution to elucidate the structure-property relationships directly from chemical data. Here we show that the performance of graph convolutional networks (GCNs) for the prediction of molecular properties can be improved by incorporating attention and gate mechanisms. The attention mechanism enables a GCN to identify atoms in different environments. The gated skip-connection further improves the GCN by updating feature maps at an appropriate rate. We demonstrate that the resulting attention- and gate-augmented GCN could extract better structural features related to a target molecular property such as solubility, polarity, synthetic accessibility and photovoltaic efficiency compared to the vanilla GCN. More interestingly, it identified two distinct parts of molecules as essential structural features for high photovoltaic efficiency, and each of them coincided with the areas of donor and acceptor orbitals for charge-transfer excitations, respectively. As a result, the new model could accurately predict molecular properties and place molecules with similar properties close to each other in a well-trained latent space, which is critical for successful molecular engineering.