Dataset Information

Memory-assisted reinforcement learning for diverse molecular de novo design.

ABSTRACT: In de novo molecular design, recurrent neural networks (RNN) have been shown to be effective methods for sampling and generating novel chemical structures. Using a technique called reinforcement learning (RL), an RNN can be tuned to target a particular section of chemical space with optimized desirable properties using a scoring function. However, ligands generated by current RL methods so far tend to have relatively low diversity, and sometimes even result in duplicate structures when optimizing towards desired properties. Here, we propose a new method to address the low diversity issue in RL for molecular design. Memory-assisted RL is an extension of the known RL, with the introduction of a so-called memory unit. As proof of concept, we applied our method to generate structures with a desired AlogP value. In a second case study, we applied our method to design ligands for the dopamine type 2 receptor and the 5-hydroxytryptamine type 1A receptor. For both receptors, a machine learning model was developed to predict whether generated molecules were active or not for the receptor. In both case studies, it was found that memory-assisted RL led to the generation of more compounds predicted to be active having higher chemical diversity, thus achieving better coverage of chemical space of known ligands compared to established RL methods.

SUBMITTER: Blaschke T

PROVIDER: S-EPMC7654024 | biostudies-literature | 2020 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Memory-assisted reinforcement learning for diverse molecular de novo design.

Blaschke Thomas T Engkvist Ola O Bajorath Jürgen J Chen Hongming H

Journal of cheminformatics 20201110 1

In de novo molecular design, recurrent neural networks (RNN) have been shown to be effective methods for sampling and generating novel chemical structures. Using a technique called reinforcement learning (RL), an RNN can be tuned to target a particular section of chemical space with optimized desirable properties using a scoring function. However, ligands generated by current RL methods so far tend to have relatively low diversity, and sometimes even result in duplicate structures when optimizin ...[more]

PMID: 33292554

Similar Datasets

Project description:Organic molecules and polymers have a broad range of applications in biomedical, chemical, and materials science fields. Traditional design approaches for organic molecules and polymers are mainly experimentally-driven, guided by experience, intuition, and conceptual insights. Though they have been successfully applied to discover many important materials, these methods are facing significant challenges due to the tremendous demand of new materials and vast design space of organic molecules and polymers. Accelerated and inverse materials design is an ideal solution to these challenges. With advancements in high-throughput computation, artificial intelligence (especially machining learning, ML), and the growth of materials databases, ML-assisted materials design is emerging as a promising tool to flourish breakthroughs in many areas of materials science and engineering. To date, using ML-assisted approaches, the quantitative structure property/activity relation for material property prediction can be established more accurately and efficiently. In addition, materials design can be revolutionized and accelerated much faster than ever, through ML-enabled molecular generation and inverse molecular design. In this perspective, we review the recent progresses in ML-guided design of organic molecules and polymers, highlight several successful examples, and examine future opportunities in biomedical, chemical, and materials science fields. We further discuss the relevant challenges to solve in order to fully realize the potential of ML-assisted materials design for organic molecules and polymers. In particular, this study summarizes publicly available materials databases, feature representations for organic molecules, open-source tools for feature generation, methods for molecular generation, and ML models for prediction of material properties, which serve as a tutorial for researchers who have little experience with ML before and want to apply ML for various applications. Last but not least, it draws insights into the current limitations of ML-guided design of organic molecules and polymers. We anticipate that ML-assisted materials design for organic molecules and polymers will be the driving force in the near future, to meet the tremendous demand of new materials with tailored properties in different fields.

Project description:In polypharmacology drugs are required to bind to multiple specific targets, for example to enhance efficacy or to reduce resistance formation. Although deep learning has achieved a breakthrough in de novo design in drug discovery, most of its applications only focus on a single drug target to generate drug-like active molecules. However, in reality drug molecules often interact with more than one target which can have desired (polypharmacology) or undesired (toxicity) effects. In a previous study we proposed a new method named DrugEx that integrates an exploration strategy into RNN-based reinforcement learning to improve the diversity of the generated molecules. Here, we extended our DrugEx algorithm with multi-objective optimization to generate drug-like molecules towards multiple targets or one specific target while avoiding off-targets (the two adenosine receptors, A1AR and A2AAR, and the potassium ion channel hERG in this study). In our model, we applied an RNN as the agent and machine learning predictors as the environment. Both the agent and the environment were pre-trained in advance and then interplayed under a reinforcement learning framework. The concept of evolutionary algorithms was merged into our method such that crossover and mutation operations were implemented by the same deep learning model as the agent. During the training loop, the agent generates a batch of SMILES-based molecules. Subsequently scores for all objectives provided by the environment are used to construct Pareto ranks of the generated molecules. For this ranking a non-dominated sorting algorithm and a Tanimoto-based crowding distance algorithm using chemical fingerprints are applied. Here, we adopted GPU acceleration to speed up the process of Pareto optimization. The final reward of each molecule is calculated based on the Pareto ranking with the ranking selection algorithm. The agent is trained under the guidance of the reward to make sure it can generate desired molecules after convergence of the training process. All in all we demonstrate generation of compounds with a diverse predicted selectivity profile towards multiple targets, offering the potential of high efficacy and low toxicity.

Project description:De novo protein design represents an attractive approach for testing and extending our understanding of metalloprotein structure and function. Here, we describe our work on the design of DF (Due Ferri or two-iron in Italian), a minimalist model for the active sites of much larger and more complex natural diiron and dimanganese proteins. In nature, diiron and dimanganese proteins protypically bind their ions in 4-Glu, 2-His environments, and they catalyze diverse reactions, ranging from hydrolysis, to O2-dependent chemistry, to decarbonylation of aldehydes. In the design of DF, the position of each atom-including the backbone, the first-shell ligands, the second-shell hydrogen-bonded groups, and the well-packed hydrophobic core-was bespoke using precise mathematical equations and chemical principles. The first member of the DF family was designed to be of minimal size and complexity and yet to display the quintessential elements required for binding the dimetal cofactor. After thoroughly characterizing its structural, dynamic, spectroscopic, and functional properties, we added additional complexity in a rational stepwise manner to achieve increasingly sophisticated catalytic functions, ultimately demonstrating substrate-gated four-electron reduction of O2 to water. We also briefly describe the extension of these studies to the design of proteins that bind nonbiological metal cofactors (a synthetic porphyrin and a tetranuclear cluster), and a Zn2+/proton antiporting membrane protein. Together these studies demonstrate a successful and generally applicable strategy for de novo metalloprotein design, which might indeed mimic the process by which primordial metalloproteins evolved. We began the design process with a highly symmetrical backbone and binding site, by using point-group symmetry to assemble the secondary structures that position the amino acid side chains required for binding. The resulting models provided a rough starting point and initial parameters for the subsequent precise design of the final protein using modern methods of computational protein design. Unless the desired site is itself symmetrical, this process requires reduction of the symmetry or lifting it altogether. Nevertheless, the initial symmetrical structure can be helpful to restrain the search space during assembly of the backbone. Finally, the methods described here should be generally applicable to the design of highly stable and robust catalysts and sensors. There is considerable potential in combining the efficiency and knowledge base associated with homogeneous metal catalysis with the programmability, biocompatibility, and versatility of proteins. While the work reported here focuses on testing and learning the principles of natural metalloproteins by designing and studying proteins one at a time, there is also considerable potential for using designed proteins that incorporate both biological and nonbiological metal ion cofactors for the evolution of novel catalysts.

Dataset Information

Memory-assisted reinforcement learning for diverse molecular de novo design.

Publications

Memory-assisted reinforcement learning for diverse molecular de novo design.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets