245,554 Hits in 4.9 sec

Exploring Chemical Space with Score-based Out-of-distribution Generation [article]

Seul Lee, Jaehyeong Jo, Sung Ju Hwang
2022 arXiv   pre-print
To generate truly novel molecules with completely different structures that may have even better properties than known molecules for de novo drug discovery, more powerful exploration in the chemical space  ...  Thus, MOOD performs conditional generation by utilizing the gradients from a property prediction network that guides the reverse-time diffusion to high-scoring regions according to multiple target properties  ...  • We experimentally demonstrate that our proposed conditional OOD molecule generation framework can generate novel molecules that are drug-like, synthesizable, and have high docking scores on five protein  ... 
arXiv:2206.07632v1 fatcat:uhosiavi4bapjgz4vzjcswwrem

Helix-MO: Sample-Efficient Molecular Optimization on Scene-Sensitive Latent Space [article]

Zhiyuan Chen, Xiaomin Fang, Zixu Hua, Yueyang Huang, Fan Wang, Hua Wu, Haifeng Wang
2022 arXiv   pre-print
Helix-MO explores the chemical space in a scene-sensitive latent space, dynamically fine-tuned by multiple kinds of learning tasks from multiple perspectives.  ...  (samples) are required to provide the optimization direction, which is a considerable expense for drug discovery.  ...  Methodology This section describes our proposed cost-effective framework (CELLS) to search for molecules that meet multiple property requirements.  ... 
arXiv:2112.00905v3 fatcat:mhbad6x6dvckxcxf6sd4aui75q

C5T5: Controllable Generation of Organic Molecules with Transformers [article]

Daniel Rothchild, Alex Tamkin, Julie Yu, Ujval Misra, Joseph Gonzalez
2021 arXiv   pre-print
We demonstrate C5T5's effectiveness on four physical properties relevant for drug discovery, showing that it learns successful and chemically intuitive strategies for altering molecules towards desired  ...  However, using generative modeling to design substances with desired properties is difficult because candidate compounds must satisfy multiple constraints, including synthetic accessibility and other metrics  ...  Separately for each target property, we excluded chemicals that had no logP value in PubChem or that were not parseable by ChemAxon's calculator.  ... 
arXiv:2108.10307v1 fatcat:yb3ihh2iofc7lnwebex3552fgm

MARS: Markov Molecular Sampling for Multi-objective Drug Discovery [article]

Yutong Xie, Chence Shi, Hao Zhou, Yuwei Yang, Weinan Zhang, Yong Yu, Lei Li
2021 arXiv   pre-print
Searching for novel molecules with desired chemical properties is crucial in drug discovery.  ...  However, it remains a big challenge to find novel and diverse compounds satisfying several properties. In this paper, we propose MARS, a method for multi-objective drug molecule discovery.  ...  However, this method mainly focuses on generating molecules that follow the observed data distribution and cannot be directly tailored for property optimization.  ... 
arXiv:2103.10432v1 fatcat:gwpidt3st5bihiczr7fz5jbaw4

Deep reinforcement learning for de novo drug design

Mariya Popova, Olexandr Isayev, Alexander Tropsha
2018 Science Advances  
The approach proposed herein can find a general use for generating targeted chemical libraries of novel compounds optimized for either a single desired property or multiple properties.  ...  We propose a novel computational strategy for de novo design of molecules with desired properties termed ReLeaSE (Reinforcement Learning for Structural Evolution).  ...  Multiple measurements were averaged. Final subsets were composed from about 20 molecules for each property.  ... 
doi:10.1126/sciadv.aap7885 pmid:30050984 pmcid:PMC6059760 fatcat:xlhibvj2uzdvxnndy6itndyhca

SyntaLinker: Automatic Fragment Linking with Deep Conditional Transformer Neural Networks

Yuyao Yang, Shuangjia Zheng, Shimin Su, Chao Zhao, Jun Xu, Hongming Chen
2020 Chemical Science  
Linking fragments to generate a focused compound library for a specific drug target is one of the challenges in fragment-based drug design (FBDD). Hereby, we propose a new program named...  ...  Fig. 5 5 Distribution of chemical properties for ChEMBL molecules and the molecules generated from SyntaLinker models.  ...  to the percentage of generated chemically valid molecules with a pair of fragments; novelty is the percentage of generated chemically valid molecules with novel linkers (not present in the training set  ... 
doi:10.1039/d0sc03126g pmid:34123096 pmcid:PMC8163338 fatcat:ontsqijhtreqrjjfyqil2xuqmm

PanGu Drug Model: Learn a Molecule Like a Human [article]

Xinyuan Lin, Chi Xu, Zhaoping Xiong, Xinfeng Zhang, Ningxi Ni, Bolin Ni, Jianlong Chang, Ruiqing Pan, Zidong Wang, Fan Yu, Hualiang Jiang, Mingyue Zheng (+1 others)
2022 bioRxiv   pre-print
In addition, PanGu molecule optimizer could optimize the chemical structures of starting molecule with improved molecular property of interest.  ...  interactions, drug-drug interactions and chemical reaction productivity), molecule generation and molecule optimization.  ...  We found that PanGu appears to be an efficient tool for molecule optimization by simply controlling the condition vectors.  ... 
doi:10.1101/2022.03.31.485886 fatcat:pm3ijwu7lrgb7cw4wadfei6fyu

Paradigm Shift: The Promise of Deep Learning in Molecular Systems Engineering and Design

Abdulelah S. Alshehri, Fengqi You
2021 Frontiers in Chemical Engineering  
We further spotlight recent advances and promising directions for several deep learning architectures, methods, and optimization platforms.  ...  the next chemical revolution.  ...  The goal of such generative models is to learn expressive continuous representations that are extended to enhance the optimization of properties and generation of novel promising molecules (Goodfellow  ... 
doi:10.3389/fceng.2021.700717 fatcat:sxsqy3ik3bf7bnzftxmlippvfy

Multiple-objective Reinforcement Learning for Inverse Design and Identification [article]

Haoran Wei, Mariefel Olarte, Garrett B. Goh
2019 arXiv   pre-print
The aim of the inverse chemical design is to develop new molecules with given optimized molecular properties or objectives.  ...  To improve the model's ability to handle a large number of molecule design objectives, we developed a Reinforcement Learning (RL) based generative framework to optimize chemical molecule generation.  ...  through curriculum-based RL for optimizing multiple objectives/constraints.  ... 
arXiv:1910.03741v1 fatcat:gf5fjxsonjdynchqshum56awru

Molecular optimization by capturing chemist's intuition using deep neural networks

Jiazhen He, Huifang You, Emil Sandström, Eva Nittinger, Esben Jannik Bjerrum, Christian Tyrchan, Werngard Czechtizky, Ola Engkvist
2021 Journal of Cheminformatics  
Typically, chemists would use their intuition to suggest chemical transformations for the starting molecule being optimized.  ...  AbstractA main challenge in drug discovery is finding molecules with a desirable balance of multiple properties.  ...  Acknowledgements Jiazhen He thanks the Molecular AI group at AstraZeneca, especially Atanas Patronov and Thierry Kogej for helpful discussions, Rocío Mercado for helpful feedback on the manuscript.  ... 
doi:10.1186/s13321-021-00497-0 pmid:33743817 fatcat:7aslbvukdzgzddjpqg2z2fdlgi

On failure modes in molecule generation and optimization

Philipp Renz, Dries Van Rompaey, Jörg Kurt Wegner, Sepp Hochreiter, Günter Klambauer
2020 Drug Discovery Today : Technologies  
These generative models are often used to optimize chemical compounds towards particular properties or a desired biological activity.  ...  There has been a wave of generative models for molecules triggered by advances in the field of Deep Learning.  ...  Goal-directed generative models for molecules are trained to produce molecules with some desired property profile, for example physical or chemical properties, bioactivities or a combination thereof.  ... 
doi:10.1016/j.ddtec.2020.09.003 pmid:33386095 fatcat:vpvaf5nn7jhu7h4khm5hcvr5w4

Scaffold-constrained molecular generation [article]

Maxime Langevin, Herve Minoux, Maximilien Levesque, Marc Bianciotto
2020 arXiv   pre-print
We directly benefit from the associated reinforcement Learning methods, allowing to design molecules optimized for different properties while exploring only the relevant chemical space.  ...  One of the major applications of generative models for drug Discovery targets the lead-optimization phase.  ...  We also compute several physico-chemical proper- We note that properties are distributed similarly for the different sets of molecules, suggesting that generated molecules populates a similar property  ... 
arXiv:2009.07778v3 fatcat:6wpxonljrnhhjbuonavmsa6jli

CogMol: Target-Specific and Selective Drug Design for COVID-19 Using Deep Generative Models [article]

Vijil Chenthamarakshan, Payel Das, Samuel C. Hoffman, Hendrik Strobelt, Inkit Padhi, Kar Wai Lim, Benjamin Hoover, Matteo Manica, Jannis Born, Teodoro Laino, Aleksandra Mojsilovic
2020 arXiv   pre-print
To generate novel and optimal drug-like molecules for unseen viral targets, CogMol leverages a protein-molecule binding affinity predictor that is trained using SMILES VAE embeddings and protein sequence  ...  In this study, we propose an end-to-end framework, named CogMol (Controlled Generation of Molecules), for designing new drug-like small molecules targeting novel viral proteins with high affinity and off-target  ...  We train multiple property predictors for controlling generation. The architecture and performance of these predictors are reported in Supp. Mat. F.  ... 
arXiv:2004.01215v2 fatcat:4braxgraqnd3vm7fm664ig7oem

Strategies of multi-objective optimization in drug discovery and development

Orazio Nicolotti, Ilenia Giangreco, Antonellina Introcaso, Francesco Leonetti, Angela Stefanachi, Angelo Carotti
2011 Expert Opinion on Drug Discovery  
The authors also discuss the potential of multi-objective strategies in controlling competing properties for absorption, distribution, metabolism and elimination/toxicity optimization.  ...  In view of this, multi-objective optimization aims to discover a set of satisfactory compromises that may in turn be used to find the global optimal solution by optimizing numerous dependent properties  ...  And in fact, the existence of multiple properties can often involve the existence of multiple solutions whose number grows up at the increase of the number of properties under optimization.  ... 
doi:10.1517/17460441.2011.588696 pmid:22646211 fatcat:enutgmi3s5hbdeiy4rg7gfymce

Generative network complex for the automated generation of druglike molecules [article]

Kaifu Gao, Duc D Nguyen, Meihua Tu, Guo-Wei Wei
2020 arXiv   pre-print
In our GNC, both multiple chemical properties and similarity scores are optimized to generate and predict drug-like molecules with desired chemical properties.  ...  In this work, we develop a generative network complex (GNC) to generate new drug-like molecules based on the multi-property optimization via the gradient descent in the latent space of an autoencoder.  ...  In this new GNC, multiple chemical properties, particularly binding affinity and similarity score, are optimized to generate new molecules with desired chemical and drug properties.  ... 
arXiv:2005.14286v1 fatcat:qbuvp5gbjzcapkmhpb7ey4cxcy
« Previous Showing results 1 — 15 out of 245,554 results