Advances and challenges in deep generative models for de novo molecule generation

Xue, Dongyu; Gong, Yukang; Yang, Zhaoyi; Chuai, Guohui; Qu, Sheng; Shen, Albert Yuh-Jer; Yu, Jing; Liu, Qi

doi:10.1002/wcms.1395

Cited by 67 publications

(70 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…and employ reinforcement learning Guimaraes et al[2017],Zhou et al [2019],Putin et al [2018a],You et al [2018],Putin et al [2018b],Yang et al [2017],Wei et al [2019],Ståhl et al [2019],Kraev [2018],,Popova et al [2019] as well as generative adversarial networksPrykhodko et al [2019] for the generative process. These methods are well-summarized by a number of recent review articles[Xue et al, 2019, Elton et al, 2019, Schwalbe-Koda and Gómez-Bombarelli, 2019, Chang, 2019, Sanchez-Lengeling and Aspuru-Guzik, 2018a. In this work we focus solely on VAEbased Bayesian optimization schemes for molecule generation and so we do not benchmark model performance against the aforementioned methods.…”

mentioning

confidence: 99%

Constrained Bayesian optimization for automatic chemical design using variational autoencoders

2020

View full text Add to dashboard Cite

Automatic Chemical Design is a framework for generating novel molecules with optimized properties. The original scheme, featuring Bayesian optimization over the latent space of a variational autoencoder, suffers from the pathology that it tends to produce invalid molecular structures. First, we demonstrate empirically that this pathology arises when the Bayesian optimization scheme queries latent points far away from the data on which the variational autoencoder has been trained. Secondly, by reformulating the search procedure as a constrained Bayesian optimization problem, we show that the effects of this pathology can be mitigated, yielding marked improvements in the validity of the generated molecules. We posit that constrained Bayesian optimization is a good approach for solving this class of training set mismatch in many generative tasks involving Bayesian optimization over the latent space of a variational autoencoder.

show abstract

mentioning

confidence: 99%

Constrained Bayesian optimization for automatic chemical design using variational autoencoders

2020

View full text Add to dashboard Cite

show abstract

“…We will also advance the development of the overarching framework ChemEco that binds the different components of our software ecosystem together and allows them to interact directly. Our long‐term vision is to enable the fully automated exploration of compound space that supports the accelerated discovery and rational design of next‐generation chemistry and materials …”

Section: Discussionmentioning

confidence: 99%

ChemML: A machine learning and informatics program package for the analysis, mining, and modeling of chemical and materials data

Haghighatlari

Vishwakarma

Altarawy

et al. 2020

WIREs Comput Mol Sci

View full text Add to dashboard Cite

ChemML is an open machine learning (ML) and informatics program suite that is designed to support and advance the data‐driven research paradigm that is currently emerging in the chemical and materials domain. ChemML allows its users to perform various data science tasks and execute ML workflows that are adapted specifically for the chemical and materials context. Key features are automation, general‐purpose utility, versatility, and user‐friendliness in order to make the application of modern data science a viable and widely accessible proposition in the broader chemistry and materials community. ChemML is also designed to facilitate methodological innovation, and it is one of the cornerstones of the software ecosystem for data‐driven in silico research. This article is categorized under: Software > Simulation Methods Computer and Information Science > Chemoinformatics Structure and Mechanism > Computational Materials Science Software > Molecular Modeling

show abstract

“…We will also advance the development of the overarching framework ChemEco that binds the different components of our software ecosystem [1] together and allows them to interact directly. Our long-term vision is to enable the fully automated exploration of compound space that supports the accelerated discovery and rational design of nextgeneration chemistry and materials [46,47].…”

Section: Discussionmentioning

confidence: 99%

ChemML: A Machine Learning and Informatics Program Package for the Analysis, Mining, and Modeling of Chemical and Materials Data

Haghighatlari

Vishwakarma²,

Altarawy³

et al. 2019

Preprint

View full text Add to dashboard Cite

<div>ChemML is an open machine learning and informatics program suite that is designed to support and advance the data-driven research paradigm that is currently emerging in the chemical and materials domain. ChemML allows its users to perform various data science tasks and execute machine learning workflows that are adapted specifically for the chemical and materials context. Key features are automation, general-purpose utility, versatility, and user-friendliness in order to make the application of modern data science a viable and widely accessible proposition in the broader chemistry and materials community. ChemML is also designed to facilitate methodological innovation, and it is one of the cornerstones of the software ecosystem for data-driven in silico research outlined in our recent publication1.</div>

show abstract

Advances and challenges in deep generative models for de novo molecule generation

Cited by 67 publications

References 32 publications

Constrained Bayesian optimization for automatic chemical design using variational autoencoders

Constrained Bayesian optimization for automatic chemical design using variational autoencoders

ChemML: A machine learning and informatics program package for the analysis, mining, and modeling of chemical and materials data

ChemML: A Machine Learning and Informatics Program Package for the Analysis, Mining, and Modeling of Chemical and Materials Data

Contact Info

Product

Resources

About