Masked graph modeling for molecule generation

Omar Mahmood, Elman Mansimov, Richard Bonneau, Kyunghyun Cho

Research output: Contribution to journalArticlepeer-review

Abstract

De novo, in-silico design of molecules is a challenging problem with applications in drug discovery and material design. We introduce a masked graph model, which learns a distribution over graphs by capturing conditional distributions over unobserved nodes (atoms) and edges (bonds) given observed ones. We train and then sample from our model by iteratively masking and replacing different parts of initialized graphs. We evaluate our approach on the QM9 and ChEMBL datasets using the GuacaMol distribution-learning benchmark. We find that validity, KL-divergence and Fréchet ChemNet Distance scores are anti-correlated with novelty, and that we can trade off between these metrics more effectively than existing models. On distributional metrics, our model outperforms previously proposed graph-based approaches and is competitive with SMILES-based approaches. Finally, we show our model generates molecules with desired values of specified properties while maintaining physiochemical similarity to the training distribution.

Original languageEnglish (US)
Article number3156
JournalNature communications
Volume12
Issue number1
DOIs
StatePublished - Dec 1 2021

ASJC Scopus subject areas

  • General Chemistry
  • General Biochemistry, Genetics and Molecular Biology
  • General Physics and Astronomy

Fingerprint

Dive into the research topics of 'Masked graph modeling for molecule generation'. Together they form a unique fingerprint.

Cite this