Molecule Attention Transformer
| Title: | Molecule Attention Transformer |
|---|---|
| Authors: | Maziarka, Łukasz, Danel, Tomasz, Mucha, Sławomir, Rataj, Krzysztof, Tabor, Jacek, Jastrzębski, Stanisław |
| Source: | Graph Representation Learning workshop and Machine Learning and the Physical Sciences workshop at NeurIPS 2019 |
| Publication Year: | 2020 |
| Collection: | Computer Science Physics (Other) Statistics |
| Subject Terms: | Computer Science - Machine Learning, Physics - Computational Physics, Statistics - Machine Learning |
| More Details: | Designing a single neural network architecture that performs competitively across a range of molecule property prediction tasks remains largely an open challenge, and its solution may unlock a widespread use of deep learning in the drug discovery industry. To move towards this goal, we propose Molecule Attention Transformer (MAT). Our key innovation is to augment the attention mechanism in Transformer using inter-atomic distances and the molecular graph structure. Experiments show that MAT performs competitively on a diverse set of molecular prediction tasks. Most importantly, with a simple self-supervised pretraining, MAT requires tuning of only a few hyperparameter values to achieve state-of-the-art performance on downstream tasks. Finally, we show that attention weights learned by MAT are interpretable from the chemical point of view. |
| Document Type: | Working Paper |
| Access URL: | http://arxiv.org/abs/2002.08264 |
| Accession Number: | edsarx.2002.08264 |
| Database: | arXiv |
| Description not available. |