Masked graph modeling for molecule generation

Generating new sensible molecular structures is a key problem in computer aided drug discovery. Here the authors propose a graph-based molecular generative model that outperforms previously proposed graph-based generative models of molecules and performs comparably to several SMILES-based models.

Bibliographic Details
Main Authors: Omar Mahmood, Elman Mansimov, Richard Bonneau, Kyunghyun Cho
Format: Article
Language:English
Published: Nature Publishing Group 2021-05-01
Series:Nature Communications
Online Access:https://doi.org/10.1038/s41467-021-23415-2
id doaj-f1a490c11d414498808c7560daa63abb
record_format Article
spelling doaj-f1a490c11d414498808c7560daa63abb2021-05-30T11:14:58ZengNature Publishing GroupNature Communications2041-17232021-05-0112111210.1038/s41467-021-23415-2Masked graph modeling for molecule generationOmar Mahmood0Elman Mansimov1Richard Bonneau2Kyunghyun Cho3Center for Data Science, New York UniversityDepartment of Computer Science, Courant Institute of Mathematical SciencesCenter for Genomics and Systems Biology, New York UniversityDepartment of Computer Science, Courant Institute of Mathematical SciencesGenerating new sensible molecular structures is a key problem in computer aided drug discovery. Here the authors propose a graph-based molecular generative model that outperforms previously proposed graph-based generative models of molecules and performs comparably to several SMILES-based models.https://doi.org/10.1038/s41467-021-23415-2
collection DOAJ
language English
format Article
sources DOAJ
author Omar Mahmood
Elman Mansimov
Richard Bonneau
Kyunghyun Cho
spellingShingle Omar Mahmood
Elman Mansimov
Richard Bonneau
Kyunghyun Cho
Masked graph modeling for molecule generation
Nature Communications
author_facet Omar Mahmood
Elman Mansimov
Richard Bonneau
Kyunghyun Cho
author_sort Omar Mahmood
title Masked graph modeling for molecule generation
title_short Masked graph modeling for molecule generation
title_full Masked graph modeling for molecule generation
title_fullStr Masked graph modeling for molecule generation
title_full_unstemmed Masked graph modeling for molecule generation
title_sort masked graph modeling for molecule generation
publisher Nature Publishing Group
series Nature Communications
issn 2041-1723
publishDate 2021-05-01
description Generating new sensible molecular structures is a key problem in computer aided drug discovery. Here the authors propose a graph-based molecular generative model that outperforms previously proposed graph-based generative models of molecules and performs comparably to several SMILES-based models.
url https://doi.org/10.1038/s41467-021-23415-2
work_keys_str_mv AT omarmahmood maskedgraphmodelingformoleculegeneration
AT elmanmansimov maskedgraphmodelingformoleculegeneration
AT richardbonneau maskedgraphmodelingformoleculegeneration
AT kyunghyuncho maskedgraphmodelingformoleculegeneration
_version_ 1721420663109976064