scholarly journals Autonomous molecule generation using reinforcement learning and docking to develop potential novel inhibitors

2020 ◽  
Vol 10 (1) ◽  
Author(s):  
Woosung Jeon ◽  
Dongsup Kim
Keyword(s):  
Reinforcement Learning ◽  
Protein Structure ◽  
Virtual Screening ◽  
Computational Method ◽  
Target Protein ◽  
Training Data ◽  
Compound Library ◽  
Lead Compounds ◽  
Large Compound ◽  
Discoidin Domain Receptor 1

AbstractWe developed a computational method named Molecule Optimization by Reinforcement Learning and Docking (MORLD) that automatically generates and optimizes lead compounds by combining reinforcement learning and docking to develop predicted novel inhibitors. This model requires only a target protein structure and directly modifies ligand structures to obtain higher predicted binding affinity for the target protein without any other training data. Using MORLD, we were able to generate potential novel inhibitors against discoidin domain receptor 1 kinase (DDR1) in less than 2 days on a moderate computer. We also demonstrated MORLD’s ability to generate predicted novel agonists for the D4 dopamine receptor (D4DR) from scratch without virtual screening on an ultra large compound library. The free web server is available at http://morld.kaist.ac.kr.

Engineering ‘Enzymelink’ for screening lead compounds to inhibit mPGES-1 while maintaining prostacyclin synthase activity

2021 ◽  
Author(s):  
Diana T Ruan ◽  
Nanhong Tang ◽  
Hironori Akasaka ◽  
Renzhong Lu ◽  
Ke-He Ruan
Keyword(s):  
Significant Degree ◽  
Hek293 Cells ◽  
Synthesis Methods ◽  
Compound Library ◽  
Lead Compounds ◽  
Cox 2 ◽  
Targeted Drug Discovery ◽  
Prostacyclin Synthase ◽  
Large Compound ◽  
Prostacyclin Synthesis

Aim: This study investigated our Enzymelinks, COX-2-10aa-mPGES-1 and COX-2-10aa-PGIS, as cellular cross-screening targets for quick identification of lead compounds to inhibit inflammatory PGE2 biosynthesis while maintaining prostacyclin synthesis. Methods: We integrated virtual and wet cross-screening using Enzymelinks to rapidly identify lead compounds from a large compound library. Results: From 380,000 compounds virtually cross-screened with the Enzymelinks, 1576 compounds were identified and used for wet cross-screening using HEK293 cells that overexpressed individual Enzymelinks as targets. The top 15 lead compounds that inhibited mPGES-1 activity were identified. The top compound that specifically inhibited inflammatory PGE2 biosynthesis alone without affecting COX-2 coupled to PGI2 synthase (PGIS) for PGI2 biosynthesis was obtained. Conclusion: Enzymelink technology could advance cyclooxygenase pathway-targeted drug discovery to a significant degree.

scholarly journals In SilicoInvestigation of Potential Pyruvate Kinase M2 Regulators from Traditional Chinese Medicine against Cancers

2014 ◽  
Vol 2014 ◽  
pp. 1-14 ◽  
Author(s):  
Kuan-Chung Chen ◽  
Kuen-Bao Chen ◽  
Hsin-Yi Chen ◽  
Calvin Yu-Chian Chen
Keyword(s):  
Virtual Screening ◽  
Drug Development ◽  
Pyruvate Kinase ◽  
Md Simulation ◽  
Docking Simulation ◽  
Target Protein ◽  
Binding Affinities ◽  
Pyruvate Kinase M2 ◽  
Lead Compounds ◽  
The Stability

A recent research in cancer research demonstrates that tumor-specific pyruvate kinase M2 (PKM2) plays an important role in chromosome segregation and mitosis progression of tumor cells. To improve the drug development of TCM compounds, we aim to identify potent TCM compounds as lead compounds of PKM2 regulators. PONDR-Fit protocol was utilized to predict the disordered disposition in the binding domain of PKM2 protein before virtual screening as the disordered structure in the protein may cause the side effect and downregulation of the possibility of ligand to bind with target protein. MD simulation was performed to validate the stability of interactions between PKM2 proteins and each ligand after virtual screening. The top TCM compounds, saussureamine C and precatorine, extracted fromLycium chinenseMill. andAbrus precatoriusL., respectively, have higher binding affinities with target protein in docking simulation than control. They have stable H-bonds with residues A:Lys311 and some other residues in both chains of PKM2 protein. Hence, we propose the TCM compounds, saussureamine C and precatorine, as potential candidates as lead compounds for further study in drug development process with the PKM2 protein against cancer.

scholarly journals Dataset Augmentation Allows Deep Learning-Based Virtual Screening To Better Generalize To Unseen Target Classes, And Highlight Important Binding Interactions

2020 ◽  
Author(s):  
Jack Scantlebury ◽  
Nathan Brown ◽  
Frank Von Delft ◽  
Charlotte M. Deane
Keyword(s):  
Deep Learning ◽  
Protein Structure ◽  
Virtual Screening ◽  
Ligand Binding ◽  
Training Data ◽  
Learning Methods ◽  
Simple Method ◽  
Physical Interactions ◽  
Account Information ◽  
Protein Models

AbstractCurrent deep learning methods for structure-based virtual screening take the structures of both the protein and the ligand as input but make little or no use of the protein structure when predicting ligand binding. Here we show how a relatively simple method of dataset augmentation forces such deep learning methods to take into account information from the protein. Models trained in this way are more generalisable (make better predictions on protein-ligand complexes from a different distribution to the training data). They also assign more meaningful importance to the protein and ligand atoms involved in binding. Overall, our results show that dataset augmentation can help deep learning based virtual screening to learn physical interactions rather than dataset biases.Graphical TOC Entry

scholarly journals In SilicoInvestigation of Potential TRAF6 Inhibitor from Traditional Chinese Medicine against Cancers

2014 ◽  
Vol 2014 ◽  
pp. 1-14 ◽  
Author(s):  
Kuan-Chung Chen ◽  
Wen-Yuan Lee ◽  
Hsin-Yi Chen ◽  
Calvin Yu-Chian Chen
Keyword(s):  
Virtual Screening ◽  
Bos Taurus ◽  
Dynamic Condition ◽  
Hypoxia Inducible Factor ◽  
Tumor Necrosis Factor Receptor ◽  
Docking Simulation ◽  
Target Protein ◽  
Lead Compounds ◽  
Hypoxia Inducible Factor 1 ◽  
The Stability

It has been indicated that tumor necrosis factor receptor-associated factor-6 (TRAF6) will upregulate the expression of hypoxia-inducible factor-1α(HIF-1α) and promote tumor angiogenesis. TRAF6 proteins can be treated as drug target proteins for a differentiation therapy against cancers. As structural disordered disposition in the protein may induce the side-effect and reduce the occupancy for ligand to bind with target protein, PONDR-Fit protocol was performed to predict the disordered disposition in TRAF6 protein before virtual screening. TCM compounds from the TCM Database@Taiwan were employed for virtual screening to identify potent compounds as lead compounds of TRAF6 inhibitor. After virtual screening, the MD simulation was performed to validate the stability of interactions between TRAF6 proteins and each ligand. The top TCM compounds, tryptophan, diiodotyrosine, and saussureamine C, extracted fromSaussurea lappaClarke,Bos taurus domesticusGmelin, andLycium chinenseMill., have higher binding affinities with target protein in docking simulation. However, the docking pose of TRAF6 protein with tryptophan is not stable under dynamic condition. For the other two TCM candidates, diiodotyrosine and saussureamine C maintain the similar docking poses under dynamic conditions. Hence, we propose the TCM compounds, diiodotyrosine and saussureamine C, as potential candidates as lead compounds for further study in drug development process with the TRAF6 protein against cancer.

scholarly journals Logic-Based Modeling and Virtual Drug Screening for the Prediction of Novel Therapeutic Targets and Combination Regimens Against E2F1-driven Melanoma Progression

2021 ◽  
Author(s):  
Nivedita Singh ◽  
Faiz M Khan ◽  
Lakshmi Bala ◽  
Julio Vera ◽  
Olaf Wolkenhauer ◽  
...  
Keyword(s):  
Virtual Screening ◽  
Regulatory Network ◽  
Epithelial Mesenchymal Transition ◽  
Therapeutic Targets ◽  
Compound Library ◽  
Mesenchymal Transition ◽  
Lead Compounds ◽  
Protein Signatures ◽  
Combination Regimens ◽  
Novel Therapeutic

Skin melanoma presents increasing prevalence and poor outcomes. Progression to aggressive stages is characterized by overexpression of the transcription factor E2F1 and activation of downstream pro-metastatic gene regulatory networks (GRNs). Appropriate therapeutic manipulation of the E2F1-governed GRNs holds potential to prevent metastasis, however these networks entail complex feedback and feedforward regulatory motifs among various regulatory layers, which challenge the characterization of drug targetablemake it difficult to identify druggable components. To this end, computational approaches such as mathematical modeling and virtual screening are important tools to unveil the dynamics of these signaling networks and comprehensively identify critical components that could be further explored as therapeutic targets. Herein, we integrated a well-established E2F1-mediated epithelial-mesenchymal transition (EMT) map with transcriptomics data from E2F1-expressing melanoma cells to reconstruct a core regulatory network underlying aggressive melanoma. Using logic-based in silico perturbation experiments of a core regulatory network, we identifiedy that simultaneous perturbation of AKT1 and MDM2 drastically reduces EMT in metastatic melanoma. Using the structures of the two protein signatures along with virtual screening of lead-like compound library available in ZINC12 database, we identified a number of lead compounds that efficiently inhibit AKT1 and MDM2 without eliciting toxicities. We propose that these compounds could be taken into account in the design of novel therapeutic strategies for the management of aggressive melanoma. were identified using virtual screening of lead-like compound library available in ZINC12 database. Subsequent high-throughput virtual screening of drug library using the structures of the two protein signatures predicted a number of lead compounds that efficiently inhibit AKT1 and MDM2 without eliciting toxicities. These can be experimentally evaluated and further considered as new anti-melanoma metastatic agents, in monotherapies or combination regimens.

scholarly journals Discovery of novel SERCA inhibitors by virtual screening of a large compound library

2011 ◽  
Vol 46 (5) ◽  
pp. 1512-1523 ◽  
Author(s):  
Christopher Elam ◽  
Michael Lape ◽  
Joel Deye ◽  
Jodie Zultowsky ◽  
David T. Stanton ◽  
...  
Keyword(s):  
Virtual Screening ◽  
Compound Library ◽  
Large Compound

scholarly journals Computational Design of Modular Robots Based on Genetic Algorithm and Reinforcement Learning

Symmetry ◽  
2021 ◽  
Vol 13 (3) ◽  
pp. 471
Author(s):  
Jai Hoon Park ◽  
Kang Hoon Lee
Keyword(s):  
Genetic Algorithm ◽  
Reinforcement Learning ◽  
Design Space ◽  
Learning Algorithm ◽  
Computational Design ◽  
Computational Method ◽  
Learning Ability ◽  
Modular Robots ◽  
Control Mechanisms ◽  
Candidate Structure

Designing novel robots that can cope with a specific task is a challenging problem because of the enormous design space that involves both morphological structures and control mechanisms. To this end, we present a computational method for automating the design of modular robots. Our method employs a genetic algorithm to evolve robotic structures as an outer optimization, and it applies a reinforcement learning algorithm to each candidate structure to train its behavior and evaluate its potential learning ability as an inner optimization. The size of the design space is reduced significantly by evolving only the robotic structure and by performing behavioral optimization using a separate training algorithm compared to that when both the structure and behavior are evolved simultaneously. Mutual dependence between evolution and learning is achieved by regarding the mean cumulative rewards of a candidate structure in the reinforcement learning as its fitness in the genetic algorithm. Therefore, our method searches for prospective robotic structures that can potentially lead to near-optimal behaviors if trained sufficiently. We demonstrate the usefulness of our method through several effective design results that were automatically generated in the process of experimenting with actual modular robotics kit.

scholarly journals Structure-Based Designing, Solvent Less Synthesis of 1,2,3,4-Tetrahydropyrimidine-5-carboxylate Derivatives: A Combined In Vitro and In Silico Screening Approach

Molecules ◽  
2021 ◽  
Vol 26 (15) ◽  
pp. 4424
Author(s):  
Uzma Arshad ◽  
Sibtain Ahmed ◽  
Nusrat Shafiq ◽  
Zaheer Ahmad ◽  
Aqsa Hassan ◽  
...  
Keyword(s):  
Virtual Screening ◽  
Catalytic Amount ◽  
Pyrimidine Derivatives ◽  
Qsar Studies ◽  
Lead Compounds ◽  
Biological Potential ◽  
Anti Cancer ◽  
Anti Oxidant ◽  
Vitro Analysis

Objective: In this study, small molecules possessing tetrahydropyrimidine derivatives have been synthesized having halogenated benzyl derivatives and carboxylate linkage. As previously reported, FDA approved halogenated pyrimidine derivatives prompted us to synthesize novel compounds in order to evaluate their biological potential. Methodology: Eight pyrimidine derivatives have been synthesized from ethyl acetoacetate, secondary amine, aromatic benzaldehyde by adding catalytic amount of CuCl2·2H2O via solvent less Grindstone multicomponent reagent method. Molecular structure reactivity and virtual screening were performed to check their biological efficacy as an anti-oxidant, anti-cancer and anti-diabetic agent. These studies were supported by in vitro analysis and QSAR studies. Results: After combined experimental and virtual screening 5c, 5g and 5e could serve as lead compounds, having low IC50 and high binding affinity.

scholarly journals Annotating proteins with generalized functional linkages

2008 ◽  
Vol 105 (46) ◽  
pp. 17700-17705 ◽  
Author(s):  
Richard Llewellyn ◽  
David S. Eisenberg
Keyword(s):  
Protein Function ◽  
Biological Process ◽  
Computational Method ◽  
Majority Voting ◽  
Target Protein ◽  
Biological Processes ◽  
Gene Products ◽  
Protein Protein Interaction ◽  
Protein Protein Interaction Networks

As genome sequencing outstrips the rate of high-quality, low-throughput biochemical and genetic experimentation, accurate annotation of protein function becomes a bottleneck in the progress of the biomolecular sciences. Most gene products are now annotated by homology, in which an experimentally determined function is applied to a similar sequence. This procedure becomes error-prone between more divergent sequences and can contaminate biomolecular databases. Here, we propose a computational method of assignment of function, termed Generalized Functional Linkages (GFL), that combines nonhomology-based methods with other types of data. Functional linkages describe pairwise relationships between proteins that work together to perform a biological task. GFL provides a Bayesian framework that improves annotation by arbitrating a competition among biological process annotations to best describe the target protein. GFL addresses the unequal strengths of functional linkages among proteins, the quality of existing annotations, and the similarity among them while incorporating available knowledge about the cellular location or individual molecular function of the target protein. We demonstrate GFL with functional linkages defined by an algorithm known as zorch that quantifies connectivity in protein–protein interaction networks. Even when using proteins linked only by indirect or high-throughput interactions, GFL predicts the biological processes of many proteins in Saccharomyces cerevisiae, improving the accuracy of annotation by 20% over majority voting.

Sign in / Sign up

Export Citation Format

Share Document