article Generative AI

GEOM-drugs revisited: toward more chemically accurate benchmarks for 3D molecule generation

Filipp Nikitin, Ian Dunn, David Ryan Koes, Olexandr Isayev

Digital Discovery Vol. 4 (11) pp. 3282–3291 2025 4 citations

Highlight

Revisiting GEOM drugs: corrected metrics and novel energy-based structural benchmark enable rigorous evaluation of 3D molecule generative models.

Abstract

Revisiting GEOM drugs: corrected metrics and novel energy-based structural benchmark enable rigorous evaluation of 3D molecule generative models.

Keywords

Cite This Paper

@article{Nikitin2025a,
  author = {Nikitin, Filipp and Dunn, Ian and Koes, David Ryan and Isayev, Olexandr},
  title = {GEOM-drugs revisited: toward more chemically accurate benchmarks for 3D molecule generation},
  year = {2025},
  journal = {Digital Discovery},
  volume = {4},
  number = {11},
  pages = {3282--3291},
  doi = {10.1039/d5dd00206k},
  url = {http://dx.doi.org/10.1039/D5DD00206K},
  publisher = {Royal Society of Chemistry (RSC)},
  keywords = {generative model},
  researchAreas = {generative-ai},
  highlight = {Revisiting GEOM drugs: corrected metrics and novel energy-based structural benchmark enable rigorous evaluation of 3D molecule generative models.},
  citations = {4}
}

Related Research Areas

Related Publications

2024

De novo molecule design towards biased properties via a deep generative framework and iterative transfer learning

Sattari K., Li D., Kalita B., Xie Y., Lighvan F. B., Isayev O., Lin J.

Digital Discovery, 3, 410–421 (2024)

Generative Ai

The RRCGAN, validated through DFT, demonstrates success in generating chemically valid molecules targeting energy gap values with 75% of the generated molecules have RE of <20% of the targeted values.

DOI
2022
cited74

Generative and reinforcement learning approaches for the automated de novo design of bioactive compounds

Korshunova M., Huang N., Capuzzi S., Radchenko D. S., Savych O., Moroz Y. S., Wells C. I., Willson T. M., Tropsha A., Isayev O.

Communications Chemistry, 5 (2022)

Generative Ai
Drug Discovery
Experiment Automation

AbstractDeep generative neural networks have been used increasingly in computational chemistry for de novo design of molecules with desired properties.

DOI
2018
cited682

Less is more: Sampling chemical space with active learning

Smith J. S., Nebgen B., Lubbers N., Isayev O., Roitberg A. E.

The Journal of Chemical Physics, 148 (2018)

Generative Ai
Experiment Automation

The development of accurate and transferable machine learning (ML) potentials for predicting molecular energetics is a challenging task.

DOI
2025

Scalable Low-Energy Molecular Conformer Generation with Quantum Mechanical Accuracy

Nikitin F., Anstine D. M., Zubatyuk R., Paliwal S. G., Isayev O.

(2025)

Ml Potentials
Generative Ai
Drug Discovery

Molecular geometry is crucial for biological activity and chemical reactivity; however, computational methods for generating 3D structures are limited by the vast scale of conformational space and the complexities of stereochemistry.

DOI
2024

Discovery of Crystallizable Organic Semiconductors with Machine Learning

Johnson H. M., Gusev F., Dull J. T., Seo Y., Priestley R. D., Isayev O., Rand B. P.

(2024)

Materials Informatics
Generative Ai

Crystalline organic semiconductors are known to have improved charge carrier mobility and exciton diffusion length in comparison to their amorphous counterparts.

DOI