Predicting Molecule Toxicity via Descriptor-based Graph Self-supervised Learning

IEEE Access. 2023. Vol. 11. P. 91842–91849.

Li X., Makarov I., Kiselev D.

Predicting molecular properties with Graph Neural Networks (GNNs) has recently drawn a lot of attention, with compound toxicity prediction being one of the biggest challenges. In cases where there is insufficient labeled molecule data, an effective approach is to pre-train GNNs on large-scale unlabeled molecular data and then fine-tune them for downstream tasks. Among pre-training strategies, node-level pre-training involves masking and predicting atom properties, while motif-based methods capture rich information in subgraphs. These approaches have shown effectiveness across various downstream tasks. However, current pre-training frameworks face two main challenges: (1) node-level auxiliary tasks do not preserve useful domain knowledge, and (2) the fusion of motif-based methods and node-level tasks is computationally extensive. To address these challenges, we propose Descriptor-based Graph Self-supervised Learning (DGSSL), a method that utilizes domain knowledge to enhance graph representation learning. Specifically, it identifies descriptor centers in molecules and encodes motif-like information as special atomic numbers in the pre-training tasks. This enables node-level self-supervised pre-training frameworks for GNNs to also capture rich information in local subgraphs. Experimental results demonstrate that our method achieves state-of-the-art performance on three toxicity-related benchmarks.

Keywords: Self-supervised learning, Graph neural networks, molecule toxicity prediction

Publication based on the results of: Models and method for analysis of unstructured data, data mining and recommender systems (2023)