Generative models for molecular discovery: Recent advances and challenges

Development of new products often relies on the discovery of novel molecules. While conventional molecular design involves using human expertise to propose, synthesize, and test new molecules, this process can be cost and time intensive, limiting the number of molecules that can be reasonably tested...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Wiley interdisciplinary reviews. Computational molecular science Jg. 12; H. 5; S. e1608 - n/a
Hauptverfasser: Bilodeau, Camille, Jin, Wengong, Jaakkola, Tommi, Barzilay, Regina, Jensen, Klavs F.
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Hoboken, USA Wiley Periodicals, Inc 01.09.2022
Wiley Subscription Services, Inc
Schlagworte:
Artificial intelligence
Benchmarks
Chemical synthesis
Computer architecture
Design
generative adversarial networks
generative models
Inverse design
Machine learning
Modelling
Molecular modelling
molecular representation
Neural networks
New products
normalizing flow models
Three dimensional models
variational autoencoders
ISSN:1759-0876, 1759-0884
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Development of new products often relies on the discovery of novel molecules. While conventional molecular design involves using human expertise to propose, synthesize, and test new molecules, this process can be cost and time intensive, limiting the number of molecules that can be reasonably tested. Generative modeling provides an alternative approach to molecular discovery by reformulating molecular design as an inverse design problem. Here, we review the recent advances in the state‐of‐the‐art of generative molecular design and discusses the considerations for integrating these models into real molecular discovery campaigns. We first review the model design choices required to develop and train a generative model including common 1D, 2D, and 3D representations of molecules and typical generative modeling neural network architectures. We then describe different problem statements for molecular discovery applications and explore the benchmarks used to evaluate models based on those problem statements. Finally, we discuss the important factors that play a role in integrating generative models into experimental workflows. Our aim is that this review will equip the reader with the information and context necessary to utilize generative modeling within their domain. This article is categorized under: Data Science > Artificial Intelligence/Machine Learning Generative modeling approaches can be used to discover novel and diverse compounds.
Bibliographie:Funding information
Edited by
Raghavan Sunoj, Associate Editor
This work was supported by Dow Chemical Company, the MIT consortium for Pharmaceutical Discovery and Synthesis, and the DARPA Accelerated Molecular Discovery program HR00111920025.
ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1759-0876
1759-0884
DOI:10.1002/wcms.1608