Article
Chemistry, Physical
Yichen Liu, Aaron T. Frank
Summary: Predicting the structure of RNA-ligand complexes is a crucial task in RNA structural biology. This study proposes a strategy to discriminate native-like poses from non-native ones using simulations and unbinding profiles analysis. The results show that characterizing the unbinding properties of individual poses can enhance pose prediction for ligands interacting with RNA aptamers.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2022)
Article
Chemistry, Multidisciplinary
Martin Buttenschoen, Garrett M. Morris, Charlotte M. Deane
Summary: In recent years, there have been significant developments in deep learning-based protein-ligand docking methods, offering great potential in terms of speed and accuracy. However, these methods often produce physically implausible molecular structures despite claims of state-of-the-art performance. It is crucial to evaluate them not only based on crystallographic RMSD but also on steric and energetic criteria. We introduce PoseBusters, a Python package that performs standard quality checks, and use it to compare different docking methods. The results show that classical docking tools currently outperform deep learning-based methods in terms of physical plausibility and generalization ability.
Article
Chemistry, Multidisciplinary
Chao Shen, Xueping Hu, Junbo Gao, Xujun Zhang, Haiyang Zhong, Zhe Wang, Lei Xu, Yu Kang, Dongsheng Cao, Tingjun Hou
Summary: Structure-based drug design relies on detailed knowledge of protein-ligand binding complexes, but accurate prediction of ligand-binding poses remains challenging. This study developed XGBoost-trained classifiers using a cross-docking dataset from the PDBbind database to discriminate binding poses, showing that specific features such as ECIF and Vina energy terms significantly impact performance, and inclusion of Vina energy terms in training can enhance model generalization.
JOURNAL OF CHEMINFORMATICS
(2021)
Article
Biochemical Research Methods
Linyuan Guo, Tian Qiu, Jianxin Wang
Summary: Protein-ligand interactions are crucial for cellular activities and drug discovery. The complexity and cost of experimental methods have led to a demand for computational approaches for deciphering protein-ligand interaction patterns. This study introduces a deep learning-based scoring function called ViTScore, which accurately identifies near-native poses from a set of poses. ViTScore shows promise as a tool for protein-ligand docking, accurately identifying potential drug targets and aiding in the design of new drugs with improved efficacy and safety.
IEEE TRANSACTIONS ON NANOBIOSCIENCE
(2023)
Article
Chemistry, Medicinal
Gen Li, Qing-Qing Dai, Guo-Bo Li
Summary: Metalloenzymes are enzymes that rely on metal ions, and comparing their active sites is crucial for enzyme design, function research, and inhibitor development. MeCOM is a method for comparing metalloenzyme active sites, which can accurately identify and compare active sites, evaluate similarity, and establish new associations between metalloenzymes.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2022)
Article
Biochemistry & Molecular Biology
Ma'mon M. Hatmal, Omar Abuyaman, Mutasem Taha
Summary: In this study, the use of multiple docked poses for machine learning-based QSAR modelling was introduced to discover potential inhibitors of the serine protease enzyme TMPRSS2. Xgboost, SVM, and RF were found to be the best machine learners with testing set accuracies reaching 90%. Three potential hits were identified by scanning known untested FDA approved drugs against TMPRSS2. Subsequent molecular dynamics simulation and covalent docking supported the new computational approach's results.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2021)
Article
Biochemistry & Molecular Biology
Guilherme M. Silva, Rosivaldo S. Borges, Kelton L. B. Santos, Leonardo B. Federico, Isaque A. G. Francischini, Suzane Q. Gomes, Mariana P. Barcelos, Rai C. Silva, Cleydson B. R. Santos, Carlos H. T. P. Silva
Summary: The mechanism of GSK-3 beta allosteric modulators is still under study and requires further investigation with the use of computational methods. The results suggest a potential for new compounds and provide new structural data for this class of modulators.
INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES
(2021)
Article
Biochemical Research Methods
Yangwei Jiang, Shi-Jie Chen
Summary: RNA molecules are critical for cellular functions at the gene expression and regulation levels, making them ideal targets for therapeutic drugs. The RLDOCK method accurately predicts RNA-ligand interactions through efficient computational procedures and is freely accessible online.
Article
Chemistry, Medicinal
Timofey V. Losev, Igor S. Gerasimov, Maria V. Panova, Alexey A. Lisov, Yana R. Abdyusheva, Polina V. Rusina, Eugenia Zaletskaya, Oleg V. Stroganov, Michael G. Medvedev, Fedor N. Novikov
Summary: Bioisosteres are molecules with different substituents but similar shapes. They are widely used in drug design to modify metabolism, bioavailability, and activity. However, predicting the affinity of bioisosteres with computational methods has been challenging due to their similarity to standard force fields. In this study, a quantum mechanical (QM)-cluster approach based on the GFN2-xTB method was developed and successfully applied to predict the biological activity change of H -> F bioisosteric replacements. The method showed superior accuracy compared to the ChemPLP scoring function and comparable to in vitro experiments, with a standard deviation of 0.60 kcal/mol.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2023)
Article
Entomology
Shuo Jin, Kun Qian, Lin He, Zan Zhang
Summary: In this study, a platform based on artificial intelligence technology was built for the batch prediction of insect-specific odorants, and a website called iORandLigandDB was developed for researchers to explore insect-specific odorants. The website provides access to the three-dimensional structures of existing odorant receptors (ORs) in insects and their docking data with related odorants. This research is important for understanding and controlling insect behavior.
Article
Chemistry, Physical
Samuel C. Gill, David L. Mobley
Summary: This study successfully samples multiple binding modes of a ligand in a single molecular dynamics simulation by developing a novel Monte Carlo move called MolDarting and coupling it with NCMC. The method shows significantly increased acceptance rates in testing phases compared to uniformly sampling internal and rotational/translational degrees of freedom in these systems.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2021)
Article
Chemistry, Medicinal
Alexandre Fassio, Laura Shub, Luca Ponzoni, Jessica McKinley, Matthew J. O'Meara, Rafaela S. Ferreira, Michael J. Keiser, Raquel C. de Melo Minardi
Summary: This paper introduces a machine learning-based drug discovery method that utilizes the LUNA toolkit to calculate and encode protein-ligand interactions into new fingerprints. The method also provides visual strategies for interpretable fingerprints. Experimental results show that this method outperforms traditional fingerprints in reproducing scores and identifying similarities. Therefore, LUNA and its interface fingerprints are promising approaches for machine learning-based drug discovery.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2022)
Article
Chemistry, Medicinal
Michel F. Sanner, Leonard Dieguez, Stefano Forli, Ewa Lis
Summary: Therapeutic peptides have gained significant interest as drugs, but peptide docking remains challenging. By using random forest classifiers, the docking efficiency of peptides can be greatly improved, paving the way for successful peptide docking rates comparable to those of small molecules.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2021)
Article
Biochemistry & Molecular Biology
Anna Tashchilova, Nadezhda Podoplelova, Alexey Sulimov, Danil Kutov, Ivan Ilin, Mikhail Panteleev, Khidmet Shikhaliev, Svetlana Medvedeva, Nadezhda Novichikhina, Andrey Potapov, Vladimir Sulimov
Summary: Complications caused by disorders in the blood coagulation system are prevalent in various medical fields, highlighting the need for the development of new and advanced drugs. This study utilized computational methods to search for potential low-molecular-weight noncovalent factor XIIa inhibitors, resulting in the identification of four compounds. Selectivity testing revealed that two of these inhibitors showed selectivity over other coagulation factors.
Article
Endocrinology & Metabolism
Dror Tobi, Eilon Krashin, Paul J. Davis, Vivian Cody, Martin Ellis, Osnat Ashur-Fabian
Summary: This study used computational simulation and structural analysis to determine the binding affinity of various thyroid hormone metabolites to the αvβ3 integrin, providing a basis for understanding the impact of these metabolites on cellular signaling in physiology and cancer.
FRONTIERS IN ENDOCRINOLOGY
(2022)