Application of Machine Learning Techniques to Predict Binding Affinity for Drug Targets: A Study of Cyclin-Dependent Kinase 2

Meta-learning-based Inductive logistic matrix completion for prediction of kinase inhibitors

Protein kinases become an important source of potential drug targets. Developing new, efficient, and safe small-molecule kinase inhibitors has become an important topic in the field of drug research and development. In contrast with traditional wet experiments which are time-consuming and expensive, machine learning-based approaches for predicting small molecule inhibitors for protein kinases are time-saving and cost-effective, which are highly desired for us. However, the issue of sample scarcity (known active and inactive compounds are usually limited for most kinases) poses a challenge to the research and development of machine learning-based kinase inhibitors' active prediction methods. To alleviate the data scarcity problem in the prediction of kinase inhibitors, in this study, we present a novel Meta-learning-based inductive logistic matrix completion method for the Prediction of Kinase Inhibitors (MetaILMC). MetaILMC adopts a meta-learning framework to learn a well-generalized model from tasks with sufficient samples, which can fast adapt to new tasks with limited samples. As MetaILMC allows the effective transfer of the prior knowledge learned from kinases with sufficient samples to kinases with a small number of samples, the proposed model can produce accurate predictions for kinases with limited data. Experimental results show that MetaILMC has excellent performance for prediction tasks of kinases with few-shot samples and is significantly superior to the state-of-the-art multi-task learning in terms of AUC, AUPR, etc., various performance metrics. Case studies also provided for two drugs to predict Kinase Inhibitory scores, further validating the proposed method's effectiveness and feasibility. SCIENTIFIC CONTRIBUTION: Considering the potential correlation between activity prediction tasks for different kinases, we propose a novel meta learning algorithm MetaILMC, which learns a prior of strong generalization capacity during meta-training from the tasks with sufficient training samples, such that it can be easily and quickly adapted to the new tasks of the kinase with scarce data during meta-testing. Thus, MetaILMC can effectively alleviate the data scarcity problem in the prediction of kinase inhibitors.

Decoding molecular mechanism of species-selective targeting of fungal versus human HSP90 using multiple replica molecular dynamics simulations and binding free energy calculations

As a highly evolutionarily conserved molecular chaperone, heat shock protein (HSP90), plays an important role in virulence traits, representing a therapeutic target for the treatment of fungal infections. The close evolutionary relationship between fungi and their human hosts poses a key challenge for the development of selective antifungal agents. In this work, molecular docking, multiple replica microsecond-based molecular dynamics (MD) simulations, and binding free energy calculations were performed to decode molecular mechanism of species-selective targeting of fungal versus human HSP90 triggered by the compound A11. MD simulations reveal that binding of compound A11 to human HSP90 nucleotide-binding domain (NBD) leads to obvious conformational changes relative to fungal HSP90 NBD. Binding free energy calculations show that the binding of compound A11 to fungal HSP90 NBD is stronger than that to human HSP90 NBD. Per residue-based free energy decomposition analysis was used to evaluate the inhibitor - residue interaction profile. The results efficiently identify the hot spot residues that play vital roles in favorable binding of compound A11 to fungal HSP90 NBD. This study is expected to provide a useful guidance for the development of selective inhibitors toward fungal HSP90.Communicated by Ramaswamy H. Sarma.

Big data and artificial intelligence (AI) methodologies for computer-aided drug design (CADD)

There have been numerous advances in the development of computational and statistical methods and applications of big data and artificial intelligence (AI) techniques for computer-aided drug design (CADD). Drug design is a costly and laborious process considering the biological complexity of diseases. To effectively and efficiently design and develop a new drug, CADD can be used to apply cutting-edge techniques to various limitations in the drug design field. Data pre-processing approaches, which clean the raw data for consistent and reproducible applications of big data and AI methods are introduced. We include the current status of the applicability of big data and AI methods to drug design areas such as the identification of binding sites in target proteins, structure-based virtual screening (SBVS), and absorption, distribution, metabolism, excretion and toxicity (ADMET) property prediction. Data pre-processing and applications of big data and AI methods enable the accurate and comprehensive analysis of massive biomedical data and the development of predictive models in the field of drug design. Understanding and analyzing biological, chemical, or pharmaceutical architectures of biomedical entities related to drug design will provide beneficial information in the biomedical big data era.

Deep Learning in Drug Design: Protein-Ligand Binding Affinity Prediction

Computational drug design relies on the calculation of binding strength between two biological counterparts especially a chemical compound, i.e., a ligand, and a protein. Predicting the affinity of protein-ligand binding with reasonable accuracy is crucial for drug discovery, and enables the optimization of compounds to achieve better interaction with their target protein. In this paper, we propose a data-driven framework named DeepAtom to accurately predict the protein-ligand binding affinity. With 3D Convolutional Neural Network (3D-CNN) architecture, DeepAtom could automatically extract binding related atomic interaction patterns from the voxelized complex structure. Compared with the other CNN based approaches, our light-weight model design effectively improves the model representational capacity, even with the limited available training data. We carried out validation experiments on the PDBbind v.2016 benchmark and the independent Astex Diverse Set. We demonstrate that the less feature engineering dependent DeepAtom approach consistently outperforms the other baseline scoring methods. We also compile and propose a new benchmark dataset to further improve the model performances. With the new dataset as training input, DeepAtom achieves Pearson's R=0.83 and RMSE=1.23 pK units on the PDBbind v.2016 core set. The promising results demonstrate that DeepAtom models can be potentially adopted in computational drug development protocols such as molecular docking and virtual screening.

Virtual screening of dipeptidyl peptidase-4 inhibitors using quantitative structure-activity relationship-based artificial intelligence and molecular docking of hit compounds

Dipeptidyl peptidase-4 (DPP-4) inhibitors are becoming an essential drug in the treatment of type 2 diabetes mellitus; however, some classes of these drugs exert side effects, including joint pain and pancreatitis. Studies suggest that these side effects might be related to secondary inhibition of DPP-8 and DPP-9. In this study, we identified DPP-4-inhibitor hit compounds selective against DPP-8 and DPP-9. We built a virtual screening workflow using a quantitative structure-activity relationship (QSAR) strategy based on artificial intelligence to allow faster screening of millions of molecules for the DPP-4 target relative to other screening methods. Five regression machine learning algorithms and four classification machine learning algorithms were applied to build virtual screening workflows, with the QSAR model applied using support vector regression (R²_pred 0.78) and the classification QSAR model using the random forest algorithm with 92.2% accuracy. Virtual screening results of > 10 million molecules obtained 2 716 hits compounds with a pIC₅₀ value of > 7.5. Additionally, molecular docking results of several potential hit compounds for DPP-4, DPP-8, and DPP-9 identified CH0002 as showing high inhibitory potential against DPP-4 and low inhibitory potential for DPP-8 and DPP-9 enzymes. These results demonstrated the effectiveness of this technique for identifying DPP-4-inhibitor hit compounds selective for DPP-4 and against DPP-8 and DPP-9 and suggest its potential efficacy for applications to discover hit compounds of other targets.

Machine-learning methods for ligand-protein molecular docking

Artificial intelligence (AI) is often presented as a new Industrial Revolution. Many domains use AI, including molecular simulation for drug discovery. In this review, we provide an overview of ligand-protein molecular docking and how machine learning (ML), especially deep learning (DL), a subset of ML, is transforming the field by tackling the associated challenges.

Computational Prediction of Binding Affinity for CDK2-ligand Complexes. A Protein Target for Cancer Drug Discovery

BACKGROUND

CDK2 participates in the control of eukaryotic cell-cycle progression. Due to the great interest in CDK2 for drug development and the relative easiness in crystallizing this enzyme, we have over 400 structural studies focused on this protein target. This structural data is the basis for the development of computational models to estimate CDK2-ligand binding affinity.

OBJECTIVE

This work focuses on the recent developments in the application of supervised machine learning modeling to develop scoring functions to predict the binding affinity of CDK2.

METHOD

We employed the structures available at the protein data bank and the ligand information accessed from the BindingDB, Binding MOAD, and PDBbind to evaluate the predictive performance of machine learning techniques combined with physical modeling used to calculate binding affinity. We compared this hybrid methodology with classical scoring functions available in docking programs.

RESULTS

Our comparative analysis of previously published models indicated that a model created using a combination of a mass-spring system and cross-validated Elastic Net to predict the binding affinity of CDK2-inhibitor complexes outperformed classical scoring functions available in AutoDock4 and AutoDock Vina.

CONCLUSION

All studies reviewed here suggest that targeted machine learning models are superior to classical scoring functions to calculate binding affinities. Specifically for CDK2, we see that the combination of physical modeling with supervised machine learning techniques exhibits improved predictive performance to calculate the protein-ligand binding affinity. These results find theoretical support in the application of the concept of scoring function space.

Recent trends in artificial intelligence-driven identification and development of anti-neurodegenerative therapeutic agents

Neurological disorders affect various aspects of life. Finding drugs for the central nervous system is a very challenging and complex task due to the involvement of the blood-brain barrier, P-glycoprotein, and the drug's high attrition rates. The availability of big data present in online databases and resources has enabled the emergence of artificial intelligence techniques including machine learning to analyze, process the data, and predict the unknown data with high efficiency. The use of these modern techniques has revolutionized the whole drug development paradigm, with an unprecedented acceleration in the central nervous system drug discovery programs. Also, the new deep learning architectures proposed in many recent works have given a better understanding of how artificial intelligence can tackle big complex problems that arose due to central nervous system disorders. Therefore, the present review provides comprehensive and up-to-date information on machine learning/artificial intelligence-triggered effort in the brain care domain. In addition, a brief overview is presented on machine learning algorithms and their uses in structure-based drug design, ligand-based drug design, ADMET prediction, de novo drug design, and drug repurposing. Lastly, we conclude by discussing the major challenges and limitations posed and how they can be tackled in the future by using these modern machine learning/artificial intelligence approaches.

Recent Advances in the Development of Thalidomide-Related Compounds as Anticancer Drugs

INTRODUCTION

Thalidomide is an old well-known drug that was first used as morning sickness relief in pregnant women before being withdrawn from the market due to its severe side effects on normal fetal development, However, over the last few decades, the interest in this old drug has been renewed because of its efficacy in several important disorders for instance, multiple myeloma, breast cancer, and HIV-related diseases due to its antiangiogenic and immunomodulatory properties. Unfortunately, even in these cases, many aftereffects as deep vein thrombosis, peripheral neuropathy, constipation, somnolence, pyrexia, pain, and teratogenicity have been reported, showing the requirement of careful and monitored use. For this reason, research efforts are geared toward the synthesis and optimization of new thalidomide analogues lacking in toxic effects to erase these limits and improve the pharmacological profile.

AIMS

This review aims to examine the state-of-the-art concerning the current studies on thalidomide and its analogues towards cancer diseases (with few hints regarding the antimicrobial activity), focusing the attention on the possible mechanisms of action involved and the lack of toxicity.

CONCLUSION

In the light of the collected data, thalidomide analogues and their ongoing optimization could lead, in the future, to the realization of a promising therapeutic alternative for cancer-fighting.

Machine Learning-Based Scoring Functions, Development and Applications with SAnDReS

BACKGROUND

Analysis of atomic coordinates of protein-ligand complexes can provide three-dimensional data to generate computational models to evaluate binding affinity and thermodynamic state functions. Application of machine learning techniques can create models to assess protein-ligand potential energy and binding affinity. These methods show superior predictive performance when compared with classical scoring functions available in docking programs.

OBJECTIVE

Our purpose here is to review the development and application of the program SAnDReS. We describe the creation of machine learning models to assess the binding affinity of protein-ligand complexes.

METHODS

SAnDReS implements machine learning methods available in the scikit-learn library. This program is available for download at https://github.com/azevedolab/sandres. SAnDReS uses crystallographic structures, binding and thermodynamic data to create targeted scoring functions.

RESULTS

Recent applications of the program SAnDReS to drug targets such as Coagulation factor Xa, cyclin-dependent kinases and HIV-1 protease were able to create targeted scoring functions to predict inhibition of these proteins. These targeted models outperform classical scoring functions.

CONCLUSION

Here, we reviewed the development of machine learning scoring functions to predict binding affinity through the application of the program SAnDReS. Our studies show the superior predictive performance of the SAnDReS-developed models when compared with classical scoring functions available in the programs such as AutoDock4, Molegro Virtual Docker and AutoDock Vina.

Alcoholism and Osteoimmunology

BACKGROUND

Chronic consumption of alcohol has an adverse effect on the skeletal system, which may lead to osteoporosis, delayed fracture healing and osteonecrosis of the femoral head. Currently, the treatment is limited, therefore, there is an urgent need to determine the underline mechanism and develop a new treatment. It is well-known that normal bone remodeling relies on the balance between osteoclast-mediated bone resorption and - mediated bone formation. Various factors can destroy the balance, including the dysfunction of the immune system. In this review, we summarized the relevant research in the alcoholic osteopenia with a focus on the abnormal osteoimmunology signals. We provided a new theoretical basis for the prevention and treatment of the alcoholic bone.

METHODS

We searched PubMed for publications from 1 January 1980 to 1 February 2020 to identify relevant and recent literature, summarizing evaluation and the prospect of alcoholic osteopenia. Detailed search terms were 'alcohol', 'alcoholic osteoporosis', 'alcoholic osteopenia' 'immune', 'osteoimmunology', 'bone remodeling', 'osteoporosis treatment' and 'osteoporosis therapy'.

RESULTS

A total of 135 papers are included in the review. About 60 papers described the mechanisms of alcohol involved in bone remodeling. Some papers were focused on the pathogenesis of alcohol on bone through osteoimmune mechanisms.

CONCLUSION

There is a complex network of signals between alcohol and bone remodeling and intercellular communication of osteoimmune may be a potential mechanism for alcoholic bone. Studying the osteoimmune mechanism is critical for drug development specific to the alcoholic bone disorder.

The First Insight Into the Supramolecular System of D,L-α-Difluoromethylornithine: A New Antiviral Perspective

Targeting the polyamine biosynthetic pathway by inhibiting ornithine decarboxylase (ODC) is a powerful approach in the fight against diverse viruses, including SARS-CoV-2. Difluoromethylornithine (DFMO, eflornithine) is the best-known inhibitor of ODC and a broad-spectrum, unique therapeutical agent. Nevertheless, its pharmacokinetic profile is not perfect, especially when large doses are required in antiviral treatment. This article presents a holistic study focusing on the molecular and supramolecular structure of DFMO and the design of its analogues toward the development of safer and more effective formulations. In this context, we provide the first deep insight into the supramolecular system of DFMO supplemented by a comprehensive, qualitative and quantitative survey of non-covalent interactions via Hirshfeld surface, molecular electrostatic potential, enrichment ratio and energy frameworks analysis visualizing 3-D topology of interactions in order to understand the differences in the cooperativity of interactions involved in the formation of either basic or large synthons (Long-range Synthon Aufbau Modules, LSAM) at the subsequent levels of well-organized supramolecular self-assembly, in comparison with the ornithine structure. In the light of the drug discovery, supramolecular studies of amino acids, essential constituents of proteins, are of prime importance. In brief, the same amino-carboxy synthons are observed in the bio-system containing DFMO. DFT calculations revealed that the biological environment changes the molecular structure of DFMO only slightly. The ADMET profile of structural modifications of DFMO and optimization of its analogue as a new promising drug via molecular docking are discussed in detail.

Targeting Mammalian Target of Rapamycin: Prospects for the Treatment of Inflammatory Bowel Diseases

Inflammatory bowel disease (IBD) is a general term for a group of chronic and progressive disorders. Several cellular and biomolecular pathways are implicated in the pathogenesis of IBD, yet the etiology is unclear. Activation of the mammalian target of rapamycin (mTOR) pathway in the intestinal epithelial cells was also shown to induce inflammation. This review focuses on the inhibition of the mTOR signaling pathway and its potential application in treating IBD. We also provide an overview of plant-derived compounds that are beneficial for the IBD management through modulation of the mTOR pathway. Data were extracted from clinical, in vitro and in vivo studies published in English between 1995 and May 2019, which were collected from PubMed, Google Scholar, Scopus and Cochrane library databases. Results of various studies implied that inhibition of the mTOR signaling pathway downregulates the inflammatory processes and cytokines involved in IBD. In this context, a number of natural products might reverse the pathological features of the disease. Furthermore, mTOR provides a novel drug target for IBD. Comprehensive clinical studies are required to confirm the efficacy of mTOR inhibitors in treating IBD.

Novel and versatile artificial intelligence algorithms for investigating possible GHSR1α and DRD1 agonists for Alzheimer's disease

Hippocampal lesions are recognized as the earliest pathological changes in Alzheimer's disease (AD). Recent researches have shown that the co-activation of growth hormone secretagogue receptor 1α (GHSR1α) and dopamine receptor D1 (DRD1) could recover the function of hippocampal synaptic and cognition. We combined traditional virtual screening technology with artificial intelligence models to screen multi-target agonists for target proteins from TCM database and a novel boost Generalized Regression Neural Network (GRNN) model was proposed in this article to improve the poor adjustability of GRNN. R-square was chosen to evaluate the accuracy of these artificial intelligent models. For the GHSR1α agonist dataset, Adaptive Boosting (AdaBoost), Linear Ridge Regression (LRR), Support Vector Machine (SVM), and boost GRNN achieved good results; the R-square of the test set of these models reached 0.900, 0.813, 0.708, and 0.802, respectively. For the DRD1 agonist dataset, Gradient Boosting (GB), Random Forest (RF), SVM, and boost GRNN achieved good results; the R-square of the test set of these models reached 0.839, 0.781, 0.763, and 0.815, respectively. According to these values of R-square, it is obvious that boost GRNN and SVM have better adaptability for different data sets and boost GRNN is more accurate than SVM. To evaluate the reliability of screening results, molecular dynamics (MD) simulation experiments were performed to make sure that candidates were docked well in the protein binding site. By analyzing the results of these artificial intelligent models and MD experiments, we suggest that 2007_17103 and 2007_13380 are the possible dual-target drugs for Alzheimer's disease (AD).

An innovative kinome platform to accelerate small-molecule inhibitor discovery and optimization from hits to leads

Kinases, accounting for 20% of the human genome, have been the focus of pharmaceutical drug discovery efforts for over three decades. Despite concerns surrounding the tractability of kinases as drug targets, it is evident that kinase drug discovery offers great potential, underscored by the US Food and Drug Administration (FDA) approval of 48 small-molecule kinase inhibitors. Despite these successes, it is challenging to identify novel kinome selective inhibitors with good pharmacokinetic/pharmacodynamic (PK/PD) properties, and resistance to kinase inhibitor treatment frequently arises. A new era of kinase drug discovery predicates the need for diverse and powerful tools to discover the next generation of kinase inhibitors. Here, we outline key tenets of the Bristol Meyers Squibb (BMS) kinase platform, to enable efficient generation of highly optimized kinase inhibitors.

Inhibiting CDK6 Activity by Quercetin Is an Attractive Strategy for Cancer Therapy

Cyclin-dependent kinase 6 (CDK6) is a potential drug target that plays an important role in the progression of different types of cancers. We performed in silico and in vitro screening of different natural compounds and found that quercetin has a high binding affinity for the CDK6 and inhibits its activity with an IC50 = 5.89 μM. Molecular docking and a 200 ns whole atom simulation of the CDK6-quercetin complex provide insights into the binding mechanism and stability of the complex. Binding parameters ascertained by fluorescence and isothermal titration calorimetry studies revealed a binding constant in the range of 107 M-1 of quercetin to the CDK6. Thermodynamic parameters associated with the formation of the CDK6-quercetin complex suggested an electrostatic interaction-driven process. The cell-based protein expression studies in the breast (MCF-7) and lung (A549) cancer cells revealed that the treatment of quercetin decreases the expression of CDK6. Quercetin also decreases the viability and colony formation potential of selected cancer cells. Moreover, quercetin induces apoptosis, by decreasing the production of reactive oxygen species and CDK6 expression. Both in silico and in vitro studies highlight the significance of quercetin for the development of anticancer leads in terms of CDK6 inhibitors.