Leveraging Classical Algorithms for Graph Neural Networks

TL;DR

Pretraining Graph Neural Networks on classical algorithms enhances their ability to predict molecular properties, demonstrating that embedding algorithmic priors improves performance on real-world graph tasks.

Contribution

This work introduces a novel approach of pretraining GNNs with classical algorithms to improve molecular property prediction accuracy.

Findings

Pretraining with classical algorithms yields consistent performance improvements.

Segments Intersect pretraining improves HIV inhibition prediction by 6%.

Dijkstra pretraining improves clinical toxicity prediction by 3%.

Abstract

Neural networks excel at processing unstructured data but often fail to generalise out-of-distribution, whereas classical algorithms guarantee correctness but lack flexibility. We explore whether pretraining Graph Neural Networks (GNNs) on classical algorithms can improve their performance on molecular property prediction tasks from the Open Graph Benchmark: ogbg-molhiv (HIV inhibition) and ogbg-molclintox (clinical toxicity). GNNs trained on 24 classical algorithms from the CLRS Algorithmic Reasoning Benchmark are used to initialise and freeze selected layers of a second GNN for molecular prediction. Compared to a randomly initialised baseline, the pretrained models achieve consistent wins or ties, with the Segments Intersect algorithm pretraining yielding a 6% absolute gain on ogbg-molhiv and Dijkstra pretraining achieving a 3% gain on ogbg-molclintox. These results demonstrate embedding classical algorithmic priors into GNNs provides useful inductive biases, boosting performance on complex, real-world graph data.