DockStream: A Docking Wrapper to Enhance De Novo Molecular Design

Guo, Jeff; Janet, Jon Paul; Bauer, Matthias R.; Nittinger, Eva; Giblin, Kathryn A.; Papadopoulos, Kostas; Voronov, Alexey; Patronov, Atanas; Engkvist, Ola; Margreitter, Christian

doi:10.26434/chemrxiv-2021-qvhml

Cited by 6 publications

(10 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We note that this analysis of scoring function suitability is out of the scope of this work but we aim to cover this in future work. On the other hand, the REINVENT strategy has been shown to maintain similar chemistry to the prior RNN [4,52,53,75]. Therefore, we visually compared some of the top molecules generated at different values of σ, shown in Additional file 1: Figure S3.…”

Section: Optimization Of Drd2 Docking Score By Augmented Hill-climb C...mentioning

confidence: 99%

“…While low sample-efficiency is not a problem for easily computed scoring functions such as property calculation, it significantly hinders the use of scoring functions requiring a significant amount of computation such as molecular docking and computer aided-synthesis planning. This is becoming increasingly important with recent growth in interest in using molecular docking scoring functions to guide de novo molecule generation [45][46][47][48][49][50][51][52][53]. This approach has shown to result in more diverse and novel compounds with a broader coverage of known active space than an equivalent QSAR model trained on known ligands [52].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Augmented Hill-Climb increases reinforcement learning efficiency for language-based de novo molecule generation

et al. 2022

View full text Add to dashboard Cite

A plethora of AI-based techniques now exists to conduct de novo molecule generation that can devise molecules conditioned towards a particular endpoint in the context of drug design. One popular approach is using reinforcement learning to update a recurrent neural network or language-based de novo molecule generator. However, reinforcement learning can be inefficient, sometimes requiring up to 105 molecules to be sampled to optimize more complex objectives, which poses a limitation when using computationally expensive scoring functions like docking or computer-aided synthesis planning models. In this work, we propose a reinforcement learning strategy called Augmented Hill-Climb based on a simple, hypothesis-driven hybrid between REINVENT and Hill-Climb that improves sample-efficiency by addressing the limitations of both currently used strategies. We compare its ability to optimize several docking tasks with REINVENT and benchmark this strategy against other commonly used reinforcement learning strategies including REINFORCE, REINVENT (version 1 and 2), Hill-Climb and best agent reminder. We find that optimization ability is improved ~ 1.5-fold and sample-efficiency is improved ~ 45-fold compared to REINVENT while still delivering appealing chemistry as output. Diversity filters were used, and their parameters were tuned to overcome observed failure modes that take advantage of certain diversity filter configurations. We find that Augmented Hill-Climb outperforms the other reinforcement learning strategies used on six tasks, especially in the early stages of training or for more difficult objectives. Lastly, we show improved performance not only on recurrent neural networks but also on a reinforcement learning stabilized transformer architecture. Overall, we show that Augmented Hill-Climb improves sample-efficiency for language-based de novo molecule generation conditioning via reinforcement learning, compared to the current state-of-the-art. This makes more computationally expensive scoring functions, such as docking, more accessible on a relevant timescale.

show abstract

Section: Optimization Of Drd2 Docking Score By Augmented Hill-climb C...mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Augmented Hill-Climb increases reinforcement learning efficiency for language-based de novo molecule generation

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Recent few years have seen deep graph learning (DGL) based on graph neural networks (GNNs) making remarkable progress in a variety of important areas, ranging from business scenarios such as finance (e.g., fraud detection and credit modeling) [28,144,9,67], e-commerce (e.g., recommendation system) [126,85], drug discovery and advanced material discovery [41,134,100,82,83]. Despite the progress, applying various DGL algorithms to real-world applications faces a Inherent Noise D train = (A + a , X + x , Y + y ) [164], [80], [87], [93], [72], [24] [101], [115] Distribution shift P train (G, Y ) = P test (G, Y )…”

Section: Trustworthy Graph Learningmentioning

confidence: 99%

“…In the past few years, DGL is becoming an active frontier of deep learning with an exponential growth of research. With advantages in modeling graph-structured data, DGL has achieved remarkable progress in many important areas, ranging from finance (e.g., fraud detection and credit modeling) [28,144,9,67], e-commerce (e.g., recommendation system) [126,85], drug discovery and advanced material discovery [41,134,100,82,83].…”

Section: Introductionmentioning

confidence: 99%

A Survey of Trustworthy Graph Learning: Reliability, Explainability, and Privacy Protection

Wu¹,

Li²,

Yu³

et al. 2022

Preprint

View full text Add to dashboard Cite

Deep graph learning has achieved remarkable progresses in both business and scientific areas ranging from finance and e-commerce, to drug and advanced material discovery. Despite these progresses, how to ensure various deep graph learning algorithms behave in a socially responsible manner and meet regulatory compliance requirements becomes an emerging problem, especially in risk-sensitive domains. Trustworthy graph learning (TwGL) aims to solve the above problems from a technical viewpoint. In contrast to conventional graph learning research which mainly cares about model performance, TwGL considers various reliability and safety aspects of the graph

show abstract

“…Here, the score function includes the interaction between the ligand and target, such as molecular docking. [8] In addition, other DL algorithms have been used to create novel molecules such as recurrent neural network (RNN), autoencoder, generative adversarial network. Despite some effort made in recent years, two limitations still hinder the practical application for de novo molecular design.…”

Section: Introductionmentioning

confidence: 99%

Widely Used and Fast De Novo Drug Design by a Protein Sequence-Based Reinforcement Learning Model

et al. 2022

Preprint

View full text Add to dashboard Cite

De novo molecular design has facilitated the exploration of large chemical space to accelerate drug discovery. Structure-based de novo method can overcome the data scarcity of active ligands by incorporating drug-target interaction into deep generative architectures. However, these strategies are bottlenecked by the small fraction of experimentally determined protein or complex structures. In addition, the cost of molecular generation is computationally expensive due to 3D representations of both molecule and protein. Here, we demonstrate a widely used and fast protein sequence-based reinforcement learning (RL) model for drug discovery. In the generative model, one of the reward components, a binding affinity predictor, is based on 1D protein sequence and molecular SMILES. As a proof of concept, the RL model was utilized to design molecules for four targets. The generated compounds showed bioactivities by the validation of both QSAR and molecular docking with experimental 3D binding pockets. We also found that the performance of generated molecules depends on the selection of data source training for the binding predictor. Furthermore, drug design for a kinase without any experimental structure, CDK20, was studied by our model. With only 1D protein sequences as input, the generated novel compounds showed favorable binding affinity based on the AlphaFold predicted structure.

show abstract

DockStream: A Docking Wrapper to Enhance De Novo Molecular Design

Cited by 6 publications

References 50 publications

Augmented Hill-Climb increases reinforcement learning efficiency for language-based de novo molecule generation

Augmented Hill-Climb increases reinforcement learning efficiency for language-based de novo molecule generation

A Survey of Trustworthy Graph Learning: Reliability, Explainability, and Privacy Protection

Widely Used and Fast De Novo Drug Design by a Protein Sequence-Based Reinforcement Learning Model

Contact Info

Product

Resources

About