Filters








73,370 Hits in 5.7 sec

Fragment-based molecular generative model with high generalization ability and synthetic accessibility [article]

Seonghwan Seo, Jaechang Lim, Woo Youn Kim
2021 arXiv   pre-print
A key feature of our model is a high generalization ability in terms of property control and fragment types.  ...  This often renders generated molecules with less correlation with target properties and low synthetic accessibility.  ...  Acknowledgements This work was supported by the Tech Incubator Program for Startup (TIPS) funded by the Ministry of SMEs and Startups (MSS, Korea) (S3031674). Notes and references  ... 
arXiv:2111.12907v1 fatcat:nosl47f4pvbizjbytwddhrzjei

CReM: chemically reasonable mutations framework for structure generation

Pavel Polishchuk
2020 Journal of Cheminformatics  
Approaches based on the deep learning models and conventional atom-based approaches may result in invalid structures and fail to address their synthetic feasibility issues.  ...  Fragment-based approaches can provide both better novelty and diversity of generated compounds but the issue of synthetic complexity of generated structure was not explicitly addressed before.  ...  Acknowledgements The author thanks Guzel Mindubaeva for the implementation and testing of some functions and Dr. Olena Mokshyna for critical reading of the manuscript.  ... 
doi:10.1186/s13321-020-00431-w pmid:33430959 fatcat:lpa33m6w7jbjvbod7xmxozqe34

Development of Natural Compound Molecular Fingerprint (NC-MFP) with the Dictionary of Natural Products (DNP) for natural product-based drug development

Myungwon Seo, Hyun Kil Shin, Yoochan Myung, Sungbo Hwang, Kyoung Tai No
2020 Journal of Cheminformatics  
NC-MFP is a scaffold-based molecular fingerprint method comprising scaffolds, scaffold-fragment connection points (SFCP), and fragments. The scaffolds of the NC-MFP have a hierarchical structure.  ...  Although several topological descriptors have been developed to describe the physical, chemical, and biological properties of organic molecules, especially synthetic compounds, and have been widely used  ...  We would like to thank Editage (https ://www.edita ge.co.kr) for editing and reviewing this manuscript for English language. 1  ... 
doi:10.1186/s13321-020-0410-3 pmid:33431009 fatcat:dsq2jlffejekvgxe3wlqhfr6ei

SYBA: Bayesian estimation of synthetic accessibility of organic compounds

Milan Voršilák, Michal Kolář, Ivan Čmelo, Daniel Svozil
2020 Journal of Cheminformatics  
Because SYBA is based merely on fragment contributions, it can be used for the analysis of the contribution of individual molecular parts to compound synthetic accessibility.  ...  SYBA (SYnthetic Bayesian Accessibility) is a fragment-based method for the rapid classification of organic compounds as easy- (ES) or hard-to-synthesize (HS).  ...  Acknowledgements Computational resources were supplied by the project "e-Infrastruktura CZ" (e-INFRA LM2018140) provided within the program Projects of Large Research, Development and Innovations Infrastructures  ... 
doi:10.1186/s13321-020-00439-2 pmid:33431015 fatcat:t3xgk7eiejf3pcwxisil4tlufe

Advances in de Novo Drug Design: From Conventional to Machine Learning Methods

Varnavas D. Mouchlis, Antreas Afantitis, Angela Serra, Michele Fratello, Anastasios G. Papadiamantis, Vassilis Aidinis, Iseult Lynch, Dario Greco, Georgia Melagraki
2021 International Journal of Molecular Sciences  
De novo drug design is a computational approach that generates novel molecular structures from atomic building blocks with no a priori relationships.  ...  Conventional methods include structure-based and ligand-based design, which depend on the properties of the active site of a biological target or its known active binders, respectively.  ...  The major challenge faced in de novo drug design is the synthetic accessibility of the generated molecular structures [9] .  ... 
doi:10.3390/ijms22041676 pmid:33562347 pmcid:PMC7915729 fatcat:zh4o44g5qzcevgvobzzcv76jv4

Structure-based de novo drug design using 3D deep generative models

Yibo Li, Jianfeng Pei, Luhua Lai
2021 Chemical Science  
Deep generative models are gaining much attention in the field of de novo molecule design. Compared to traditional methods, deep generative models can be trained in a fully data-driven way...  ...  Fig. 4 b shows one of the molecules with high bioactivity (compound 5, IC 50 =0.128µM), along with its drug-likeness (QED), synthetical accessibility (SAscore), and the docking score.  ...  Using L-Net for structure-based molecule design Compared to previous 3D molecular generative models 32,33 , a major advantage of L-Net is its ability to generate 3D molecular structure end-to-end.  ... 
doi:10.1039/d1sc04444c pmid:34760151 pmcid:PMC8549794 fatcat:qejosqvl3vdw3ds3x2p2oa5ohy

Molecular Design in Synthetically Accessible Chemical Space via Deep Reinforcement Learning [article]

Julien Horwood, Emmanuel Noutahi
2020 arXiv   pre-print
Despite recent progress, we argue that existing generative methods are limited in their ability to favourably shift the distributions of molecular properties during optimization.  ...  We instead propose a novel Reinforcement Learning framework for molecular design in which an agent learns to directly optimize through a space of synthetically-accessible drug-like molecules.  ...  The methods and algorithms presented here were developed at InVivo AI.  ... 
arXiv:2004.14308v2 fatcat:dbmnzvy3mfdl5np7hlnwsz3ama

Expanding medicinal chemistry space

Andy Barker, Jason G. Kettle, Thorsten Nowak, J. Elizabeth Pease
2013 Drug Discovery Today  
However, many high value biological targets lie outside this chemical space, and an ability to access such 'intractable' targets not amenable to traditional small molecule intervention would expand treatment  ...  We review these new approaches and their ability to provide the novel leads needed to tackle a new generation of biological targets. Reviews POST SCREEN  ...  Acknowledgements The authors gratefully acknowledge Dr Peter Greasley for the data used in Fig. 1 and Dr Mike Waring for helpful discussions during the preparation of this manuscript.  ... 
doi:10.1016/j.drudis.2012.10.008 pmid:23117010 fatcat:v7vgbgticzayheveusbiadyq4a

Molecular Generation with Recurrent Neural Networks (RNNs) [article]

Esben Jannik Bjerrum, Richard Threlfall
2017 arXiv   pre-print
This discrepancy has led to an interest in generating virtual libraries using hand crafted chemical rules and fragment based methods to cover a larger area of chemical space and generate chemical libraries  ...  The networks can to a high extent generate novel, but chemically sensible molecules.  ...  Figure 3 : 3 Examples of molecules generated from the model trained on the dataset with molecular fragments (p12). Figure 4 : 4 Molecular sampling error at different sampling temperature.  ... 
arXiv:1705.04612v2 fatcat:ps4etucbjbcybhuep5nczopdr4

Molecular Design in Synthetically Accessible Chemical Space via Deep Reinforcement Learning

Julien Horwood, Emmanuel Noutahi
2020 ACS Omega  
Despite recent progress, we argue that existing generative methods are limited in their ability to favorably shift the distributions of molecular properties during optimization.  ...  We instead propose a novel Reinforcement Learning framework for molecular design in which an agent learns to directly optimize through a space of synthetically accessible drug-like molecules.  ...  To overcome these limitations, Button et al. 9 proposed a hybrid rule-based and machine learning approach in which molecules are assembled from fragments under synthetic accessibility constraints in  ... 
doi:10.1021/acsomega.0c04153 pmid:33403260 pmcid:PMC7774092 fatcat:baqmf7gtp5ewlbqh6r4cqbun4m

Can we predict materials that can be synthesised?

Filip T. Szczypiński, Steven Bennett, Kim E. Jelfs
2021 Chemical Science  
Materials discovery is a crucial yet experimentally slow and wasteful process. We discuss how discovery can be accelerated by focusing on making predictions that are synthetically realisable.  ...  Acknowledgements We thank Dr Rebecca Greenaway Notes and references  ...  Currently, Filip is working with Dr Kim Jelfs at Imperial College London on the modelling of new porous organic molecular materials, where he combines his synthetic experience with computational tools  ... 
doi:10.1039/d0sc04321d pmid:34163850 pmcid:PMC8178993 fatcat:mw5bb3kodra6fmqcswhbdnlqae

Predicting reaction performance in C–N cross-coupling using machine learning

Derek T. Ahneman, Jesús G. Estrada, Shishi Lin, Spencer D. Dreher, Abigail G. Doyle
2018 Science  
We created scripts to compute and extract atomic, molecular, and vibrational descriptors for the components of a palladium-catalyzed Buchwald-Hartwig cross-coupling of aryl halides with 4-methylaniline  ...  The random forest model was also successfully applied to sparse training sets and out-of-sample prediction, suggesting its value in facilitating adoption of synthetic methodology.  ...  We have taken steps to automate reaction parameterization and modeling with the aim of making this tool accessible to the synthetic chemistry community.  ... 
doi:10.1126/science.aar5169 pmid:29449509 fatcat:55n7nzte55bgzixsgp3fculaxy

Profiling and analysis of chemical compounds using pointwise mutual information

I. Čmelo, M. Voršilák, D. Svozil
2021 Journal of Cheminformatics  
However, ZRFT ability to distinguish between ES and HS compounds is surprisingly high considering that while SYBA, SAScore and random forest are dedicated SA models, ZRFT is a generic measurement that  ...  ZRFT value distributions are compared with these of SYBA and SAScore.  ...  Acknowledgements Computational resources were supplied by the project "e-Infrastruktura CZ" (e-INFRA LM2018140) provided within the program Projects of Large Research, Development and Innovations Infrastructures  ... 
doi:10.1186/s13321-020-00483-y pmid:33423694 fatcat:zqutq532wjabphefk4sjht2jpm

Chemistry42: An AI-based platform for de novo molecular design [article]

Yan A. Ivanenkov, Alex Zhebrak, Dmitry Bezrukov, Bogdan Zagribelnyy, Vladimir Aladinskiy, Daniil Polykovskiy, Evgeny Putin, Petrina Kamya, Alexander Aliper, Alex Zhavoronkov
2021 arXiv   pre-print
Chemistry42 is unique in its ability to generate novel molecular structures with predefined properties validated through in vitro and in vivo studies.  ...  Chemistry42 is a software platform for de novo small molecule design that integrates Artificial Intelligence (AI) techniques with computational and medicinal chemistry methods.  ...  Acknowledgements The authors gratefully acknowledge the valuable comments and suggestions made by Dr. Jiye Shi from UCB Pharma (Slough, UK).  ... 
arXiv:2101.09050v1 fatcat:sxodprcm6fdq3phmll6hz6lwxq

Synthetic Biology Advanced Natural Product Discovery

Junyang Wang, Jens Nielsen, Zihe Liu
2021 Metabolites  
With the rapid development of synthetic biology, advanced genome mining and engineering strategies have been reported and they provide new opportunities for discovery of natural products.  ...  With the rapid development of DNA sequencing technology and bioinformatics, a large number of putative biosynthetic gene clusters have been reported.  ...  The new generation of genome editing tools based on the CRISPR-Cas technology has the advantages of high efficiency, fast operation and high fidelity.  ... 
doi:10.3390/metabo11110785 pmid:34822443 pmcid:PMC8617713 fatcat:x35ve3lcv5hidbborm3begugb4
« Previous Showing results 1 — 15 out of 73,370 results