AI tools such as ChatGPT can provide responses to queries on any topic, but can such large language models accurately ‘write’ molecules as output to our specification? Results now show that models trained on general text can be tweaked with small amounts of chemical data to predict molecular properties, or to design molecules based on a target feature.
This is a preview of subscription content, log in via an institution to check access.
References
Jablonka, K. M. et al. Nat. Mach. Intell. 6, 161–169 (2024).
Brown, T. B. et al. In Advances in Neural Information Processing Systems 33 (NeurIPS 2020).
Hocky, G. M. & White, A. D. Digital Discovery 1, 79–83 (2022).
Lála, J. et al. Preprint at https://doi.org/10.48550/arXiv.2312.07559 (2023).
Bran, A. M. et al. Preprint at https://doi.org/10.48550/arXiv.2304.05376 (2023).
Boiko, D. A. et al. Nature 624, 570–578 (2023).
White, A. D. et al. Digital Discovery 2, 368–376 (2023).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Competing interests
The author declares no competing interests.
Rights and permissions
About this article
Cite this article
Hocky, G.M. Connecting molecular properties with plain language. Nat Mach Intell 6, 249–250 (2024). https://doi.org/10.1038/s42256-024-00812-y
Published:
Issue Date:
DOI: https://doi.org/10.1038/s42256-024-00812-y
- Springer Nature Limited