GuacaMol: Benchmarking Models for de Novo Molecular Design

Brown, Nathan; Fiscato, Marco; Segler, Marwin; Vaucher, Alain C.

doi:10.1021/acs.jcim.8b00839

Cited by 632 publications

(967 citation statements)

References 95 publications

Supporting

Mentioning

883

Contrasting

Unclassified

Order By: Relevance

“…There are an increasing number of algorithms of these two categories proposed in recent years and a small number of studies that benchmark these algorithms in terms of their ability to generate novel, optimal molecules. 26,27 We categorize the approaches one might take to ensure that computationally designed molecules are able to be synthesized in Figure 1. These represent combinations of (i) a database of known or enumerated compounds, (ii) an evaluator, which estimates the properties we are trying to optimize, (iii) a generator function, which can propose new candidate molecules, (iv) a synthesizability oracle that determines whether it is straightforward to synthesize a given molecule, and/or (v) a heuristic synthesizability estimator that provides a computationally-inexpensive scalar measure of synthesizability.…”

Section: Introductionmentioning

confidence: 99%

The Synthesizability of Molecules Proposed by Generative Models

Gao

Coley

2020

J. Chem. Inf. Model.

247

297

View full text Add to dashboard Cite

The discovery of functional molecules is an expensive and time-consuming process, exemplified by the rising costs of small molecule therapeutic discovery. One class of techniques of growing interest for early-stage drug discovery is de novo molecular generation and optimization, catalyzed by the development of new deep learning approaches. 1 These techniques can suggest novel molecular structures intended to maximize a multi-objective function, e.g., suitability as a therapeutic against a particular target, 2 without relying on brute-force exploration of a chemical space. 3 However, the utility of these approaches is stymied by ignorance of synthesizability. To highlight the severity of this issue, we use a data-driven computer-aided synthesis planning program 4 to quantify how often molecules proposed by state-of-the-art generative models cannot be readily synthesized. Our analysis demonstrates that there are several tasks for which these models generate unrealistic molecular structures despite performing well on popular quantitative benchmarks. Synthetic complexity heuristics can successfully bias generation toward synthetically-tractable chemical space, although doing so 1 arXiv:2002.07007v1 [q-bio.QM] 17 Feb 2020 necessarily detracts from the primary objective. This analysis suggests that to improve the utility of these models in real discovery workflows, new algorithm development is warranted.

show abstract

Section: Introductionmentioning

confidence: 99%

The Synthesizability of Molecules Proposed by Generative Models

Gao

Coley

2020

J. Chem. Inf. Model.

247

297

View full text Add to dashboard Cite

show abstract

“…We assessed the generated molecules with a range of 2D and 3D metrics. As is standard in the assessment of models for molecule generation, 37 we first checked the generated molecules for validity, uniqueness, and novelty. We then determined if the generated linkers were consistent with the 2D property filters used to produce the training set.…”

Section: Methodsmentioning

confidence: 99%

Deep Generative Models for 3D Compound Design

Imrie

Bradley

Schaar

et al. 2019

Preprint

View full text Add to dashboard Cite

Rational compound design remains a challenging problem for both computational methods and medicinal chemists. Computational generative methods have begun to show promising results for the design problem. However, they have not yet used the power of 3D structural information. We have developed a novel graph-based deep generative model that combines state-of-the-art machine learning techniques with structural knowledge. Our method ("DeLinker") takes two fragments or partial structures and designs a molecule incorporating both. The generation process is protein context dependent, utilising the relative distance and orientation between the partial structures. This 3D information is vital to successful compound design, and we demonstrate its impact on the generation process and the limitations of omitting such information. In a large scale evaluation, DeLinker designed 60% more molecules with high 3D similarity to the original molecule than a database baseline. When considering the more relevant problem of longer linkers with at least five atoms, the outperformance increased to 200%. We demonstrate the effectiveness and applicability of this approach on a diverse range of design problems: fragment linking, scaffold hopping, and proteolysis tar-geting chimera (PROTAC) design. As far as we are aware, this is the first molecular generative model to incorporate 3D structural information directly in the design process. Code is available at https://github.com/oxpig/DeLinker.

show abstract

“…Modern machine learning tools have become impactful, perhaps even essential, in the early stages of the drug design process. Today, deep learning models excel at predicting various chemical properties [1], and deep generative models in conjunction with reinforcement learning are able to efficiently search molecular space for molecules that can optimize several chemical properties simultaneously [2].…”

Section: Introductionmentioning

confidence: 99%

A Novel Machine Learning Approach Uncovers New and Distinctive Inhibitors for Cyclin-Dependent Kinase 9

Aßmann¹,

Bal²,

Craig³

et al. 2020

Preprint

View full text Add to dashboard Cite

We present a novel combination of generative and predictive machine learning models for discovering unique protein inhibitors. The new method is assessed on its ability to generate unique inhibitors for the cancer associated protein kinase, CDK9. We validate our method by performing biochemical assays, attaining a hit rate of more than 10%, demonstrating the method to be a notable improvement upon a more standard, and somewhat naive approach. Moreover, we imposed the additional challenge of finding inhibitors that are readily synthesized. Importantly, two new inhibitors are found, with one being distinct from reported CDK9 inhibitors. We discuss the results in the context of modern machine learning principles and the desire expressed by the rational drug design community to secure molecules that are structurally different, yet with high binding affinities, to structurally determined protein targets.

show abstract

GuacaMol: Benchmarking Models for de Novo Molecular Design

Cited by 632 publications

References 95 publications

The Synthesizability of Molecules Proposed by Generative Models

The Synthesizability of Molecules Proposed by Generative Models

Deep Generative Models for 3D Compound Design

A Novel Machine Learning Approach Uncovers New and Distinctive Inhibitors for Cyclin-Dependent Kinase 9

Contact Info

Product

Resources

About