Translation between Molecules and Natural Language

Abstract

We present $\text{MolT5}$ $-$ a self-supervised learning framework for pretraining models on a vast amount of unlabeled natural language text and molecule strings. $\text{MolT5}$ allows for new, useful, and challenging analogs of traditional vision-language tasks, such as molecule captioning and text-based de novo molecule generation (altogether: translation between molecules and language), which we explore for the first time. Since $\text{MolT5}$ pretrains models on single-modal data, it helps overcome the chemistry domain shortcoming of data scarcity. Furthermore, we consider several metrics, including a new cross-modal embedding-based metric, to evaluate the tasks of molecule captioning and text-based molecule generation. Our results show that $\text{MolT5}$-based models are able to generate outputs, both molecules and captions, which in many cases are high quality.

Accepted at EMNLP 2022. Data and code can be found on [Github](https://github.com/blender-nlp/MolT5)