Feature selection in molecular graph neural networks based on quantum chemical approaches

19 January 2023, Version 3
This content is a preprint and has not undergone peer review at the time of posting.


Feature selection is an important topic that has been widely studied in data science. Recently, graph neural networks (GNNs) and graph convolutional networks (GCNs) have also been employed in chemistry. To enhance the performance characteristics of the GNN and GCN in the field of chemistry, the feature selection should also be discussed in detail from the chemistry viewpoint. Thus, this study proposes a new feature in molecular GNNs based on the quantum chemical approaches and discusses the accuracy, overcorrelation between features, and interpretability. From the overcorrelation and accuracy, the important graph convolution (IGC) with molecular-atomic properties (MAP) proposed herein showed good performance. Moreover, the integrated gradients analysis showed that the machine learning model with the IGC(MAP) explained the prediction outputs reasonably.


graph neural network
feature selection
graph convolutional network

Supplementary materials

Supporting information
Details of the machine learning process and RMSEs of the pKa values obtained by IGC and two types of atom features (IAP and MAP) with a different number of convolution layers.


Comments are not moderated before they are posted, but they can be removed by the site moderators if they are found to be in contravention of our Commenting Policy [opens in a new tab] - please read this policy before you post. Comments should be used for scholarly discussion of the content in question. You can find more information about how to use the commenting feature here [opens in a new tab] .
This site is protected by reCAPTCHA and the Google Privacy Policy [opens in a new tab] and Terms of Service [opens in a new tab] apply.