Improving structural plausibility in diffusion-based 3D molecule generation <i>via</i> property-conditioned training with distorted molecules
Conditional diffusion models improve molecular generation by learning from both high- and low-quality structures, using quality-labeled distorted molecules to enhance plausibility and enable selective sampling.
Authors
- Charlotte M. Deane (ORCID: https://orcid.org/0000-0003-1388-2252)
- Vijil Chenthamarakshan (ORCID: https://orcid.org/0000-0001-7830-5777)
- Payel Das (ORCID: https://orcid.org/0000-0002-7288-0516)
- Lucy Vost (ORCID: https://orcid.org/0000-0002-3194-0172)
Institutions
- University of Oxford (GB)
Publication Details
- Journal
- Digital Discovery
- Published
- 2025-01-01
- DOI
- https://doi.org/10.1039/d4dd00331d
- Citations
- 4
- Primary Topic
- Microfluidic and Capillary Electrophoresis Applications
- Type
- article
- Field-Weighted Citation Impact
- 2.70