SmilesT5: Domain-specific pretraining for molecular language models
This work addresses molecular property prediction for drug discovery, offering incremental improvements in efficiency and performance over existing methods.
The paper tackled molecular property prediction by introducing domain-specific pretraining tasks for SMILES-based language models, achieving improved performance on six classification benchmarks and demonstrating enhanced data and computational efficiency.
Molecular property prediction is an increasingly critical task within drug discovery and development. Typically, neural networks can learn molecular properties using graph-based, language-based or feature-based methods. Recent advances in natural language processing have highlighted the capabilities of neural networks to learn complex human language using masked language modelling. These approaches to training large transformer-based deep learning models have also been used to learn the language of molecules, as represented by simplified molecular-input line-entry system (SMILES) strings. Here, we present novel domain-specific text-to-text pretraining tasks that yield improved performance in six classification-based molecular property prediction benchmarks, relative to both traditional likelihood-based training and previously proposed fine-tuning tasks. Through ablation studies, we show that data and computational efficiency can be improved by using these domain-specific pretraining tasks. Finally, the pretrained embeddings from the model can be used as fixed inputs into a downstream machine learning classifier and yield comparable performance to finetuning but with much lower computational overhead.