SAnDReS 2.0: Development of machine-learning models to explore the scoring function space.

Classical scoring functions may exhibit low accuracy in determining ligand binding affinity for proteins. The availability of both protein-ligand structures and affinity data make it possible to develop machine-learning models focused on specific protein systems with superior predictive performance. Here, we report a new methodology named SAnDReS that combines AutoDock Vina 1.2 with 54 regression methods available in Scikit-Learn to calculate binding affinity based on protein-ligand structures. This approach allows exploration of the scoring function space. SAnDReS generates machine-learning models based on crystal, docked, and AlphaFold-generated structures. As a proof of concept, we examine the performance of SAnDReS-generated models in three case studies. For all three cases, our models outperformed classical scoring functions. Also, SAnDReS-generated models showed predictive performance close to or better than other machine-learning models such as KDEEP, CSM-lig, and ΔVinaRF20. SAnDReS 2.0 is available to download at https://github.com/azevedolab/sandres.

Exploring Scoring Function Space: Developing Computational Models for Drug Discovery.

BACKGROUND: The idea of scoring function space established a systems-level approach to address the development of models to predict the affinity of drug molecules by those interested in drug discovery. OBJECTIVE: Our goal here is to review the concept of scoring function space and how to explore it to develop machine learning models to address protein-ligand binding affinity. METHOD: We searched the articles available in PubMed related to the scoring function space. We also utilized crystallographic structures found in the protein data bank (PDB) to represent the protein space. RESULTS: The application of systems-level approaches to address receptor-drug interactions allows us to have a holistic view of the process of drug discovery. The scoring function space adds flexibility to the process since it makes it possible to see drug discovery as a relationship involving mathematical spaces. CONCLUSION: The application of the concept of scoring function space has provided us with an integrated view of drug discovery methods. This concept is useful during drug discovery, where we see the process as a computational search of the scoring function space to find an adequate model to predict receptor-drug binding affinity.

Protein-Ligand Docking Simulations with AutoDock4 Focused on the Main Protease of SARS-CoV-2.

BACKGROUND: The main protease of SARS-CoV-2 (Mpro) is one of the targets identified in SARS-CoV-2, the causative agent of COVID-19. The application of X-ray diffraction crystallography made available the three-dimensional structure of this protein target in complex with ligands, which paved the way for docking studies. OBJECTIVE: Our goal here is to review recent efforts in the application of docking simulations to identify inhibitors of the Mpro using the program AutoDock4. METHODS: We searched PubMed to identify studies that applied AutoDock4 for docking against this protein target. We used the structures available for Mpro to analyze intermolecular interactions and reviewed the methods used to search for inhibitors. RESULTS: The application of docking against the structures available for the Mpro found ligands with an estimated inhibition in the nanomolar range. Such computational approaches focused on the crystal structures revealed potential inhibitors of Mpro that might exhibit pharmacological activity against SARS-CoV-2. Nevertheless, most of these studies lack the proper validation of the docking protocol. Also, they all ignored the potential use of machine learning to predict affinity. CONCLUSION: The combination of structural data with computational approaches opened the possibility to accelerate the search for drugs to treat COVID-19. Several studies used AutoDock4 to search for inhibitors of Mpro. Most of them did not employ a validated docking protocol, which lends support to critics of their computational methodology. Furthermore, one of these studies reported the binding of chloroquine and hydroxychloroquine to Mpro. This study ignores the scientific evidence against the use of these antimalarial drugs to treat COVID-19.

Electrostatic Potential Energy in Protein-Drug Complexes.

BACKGROUND: Electrostatic interactions are one of the forces guiding the binding of molecules to proteins. The assessment of this interaction through computational approaches makes it possible to evaluate the energy of protein-drug complexes. OBJECTIVE: Our purpose here is to review some of the methods used to calculate the electrostatic energy of protein-drug complexes and explore the capacity of these approaches for the generation of new computational tools for drug discovery using the abstraction of scoring function space. METHODS: Here, we present an overview of the AutoDock4 semi-empirical scoring function used to calculate binding affinity for protein-drug complexes. We focus our attention on electrostatic interactions and how to explore recently published results to increase the predictive performance of the computational models to estimate the energetics of protein- drug interactions. Public data available at Binding MOAD, BindingDB, and PDBbind were used to review the predictive performance of different approaches to predict binding affinity. RESULTS: A comprehensive outline of the scoring function used to evaluate potential energy available in docking programs is presented. Recent developments of computational models to predict protein-drug energetics were able to create targeted-scoring functions to predict binding to these proteins. These targeted models outperform classical scoring functions and highlight the importance of electrostatic interactions in the definition of the binding. CONCLUSION: Here, we reviewed the development of scoring functions to predict binding affinity through the application of a semi-empirical free energy scoring function. Our studies show the superior predictive performance of machine learning models when compared with classical scoring functions and the importance of electrostatic interactions for binding affinity.

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.

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

BACKGROUND: The elucidation of the structure of cyclin-dependent kinase 2 (CDK2) made it possible to develop targeted scoring functions for virtual screening aimed to identify new inhibitors for this enzyme. CDK2 is a protein target for the development of drugs intended to modulate cellcycle progression and control. Such drugs have potential anticancer activities. OBJECTIVE: Our goal here is to review recent applications of machine learning methods to predict ligand- binding affinity for protein targets. To assess the predictive performance of classical scoring functions and targeted scoring functions, we focused our analysis on CDK2 structures. METHODS: We have experimental structural data for hundreds of binary complexes of CDK2 with different ligands, many of them with inhibition constant information. We investigate here computational methods to calculate the binding affinity of CDK2 through classical scoring functions and machine- learning models. RESULTS: Analysis of the predictive performance of classical scoring functions available in docking programs such as Molegro Virtual Docker, AutoDock4, and Autodock Vina indicated that these methods failed to predict binding affinity with significant correlation with experimental data. Targeted scoring functions developed through supervised machine learning techniques showed a significant correlation with experimental data. CONCLUSION: Here, we described the application of supervised machine learning techniques to generate a scoring function to predict binding affinity. Machine learning models showed superior predictive performance when compared with classical scoring functions. Analysis of the computational models obtained through machine learning could capture essential structural features responsible for binding affinity against CDK2.

Structural Basis for Inhibition of Enoyl-[Acyl Carrier Protein] Reductase (InhA) from Mycobacterium tuberculosis.

BACKGROUND: The enzyme trans-enoyl-[acyl carrier protein] reductase (InhA) is a central protein for the development of antitubercular drugs. This enzyme is the target for the pro-drug isoniazid, which is catalyzed by the enzyme catalase-peroxidase (KatG) to become active. OBJECTIVE: Our goal here is to review the studies on InhA, starting with general aspects and focusing on the recent structural studies, with emphasis on the crystallographic structures of complexes involving InhA and inhibitors. METHOD: We start with a literature review, and then we describe recent studies on InhA crystallographic structures. We use this structural information to depict protein-ligand interactions. We also analyze the structural basis for inhibition of InhA. Furthermore, we describe the application of computational methods to predict binding affinity based on the crystallographic position of the ligands. RESULTS: Analysis of the structures in complex with inhibitors revealed the critical residues responsible for the specificity against InhA. Most of the intermolecular interactions involve the hydrophobic residues with two exceptions, the residues Ser 94 and Tyr 158. Examination of the interactions has shown that many of the key residues for inhibitor binding were found in mutations of the InhA gene in the isoniazid-resistant Mycobacterium tuberculosis. Computational prediction of the binding affinity for InhA has indicated a moderate uphill relationship with experimental values. CONCLUSION: Analysis of the structures involving InhA inhibitors shows that small modifications on these molecules could modulate their inhibition, which may be used to design novel antitubercular drugs specific for multidrug-resistant strains.

Taba: A Tool to Analyze the Binding Affinity.

Evaluation of ligand-binding affinity using the atomic coordinates of a protein-ligand complex is a challenge from the computational point of view. The availability of crystallographic structures of complexes with binding affinity data opens the possibility to create machine-learning models targeted to a specific protein system. Here, we describe a new methodology that combines a mass-spring system approach with supervised machine-learning techniques to predict the binding affinity of protein-ligand complexes. The combination of these techniques allows exploring the scoring function space, generating a model targeted to a protein system of interest. The new model shows superior predictive performance when compared with classical scoring functions implemented in the programs Molegro Virtual Docker, AutoDock4, and AutoDock Vina. We implemented this methodology in a new program named Taba. Taba is implemented in Python and available to download under the GNU license at https://github.com/azevedolab/taba. © 2019 Wiley Periodicals, Inc.

How Docking Programs Work.

Protein-ligand docking simulations are of central interest for computer-aided drug design. Docking is also of pivotal importance to understand the structural basis for protein-ligand binding affinity. In the last decades, we have seen an explosion in the number of three-dimensional structures of protein-ligand complexes available at the Protein Data Bank. These structures gave further support for the development and validation of in silico approaches to address the binding of small molecules to proteins. As a result, we have now dozens of open source programs and web servers to carry out molecular docking simulations. The development of the docking programs and the success of such simulations called the attention of a broad spectrum of researchers not necessarily familiar with computer simulations. In this scenario, it is essential for those involved in experimental studies of protein-ligand interactions and biophysical techniques to have a glimpse of the basics of the protein-ligand docking simulations. Applications of protein-ligand docking simulations to drug development and discovery were able to identify hits, inhibitors, and even drugs. In the present chapter, we cover the fundamental ideas behind protein-ligand docking programs for non-specialists, which may benefit from such knowledge when studying molecular recognition mechanism.

SAnDReS: A Computational Tool for Docking.

Since the early 1980s, we have witnessed considerable progress in the development and application of docking programs to assess protein-ligand interactions. Most of these applications had as a goal the identification of potential new binders to protein targets. Another remarkable progress is taking place in the determination of the structures of protein-ligand complexes, mostly using X-ray diffraction crystallography. Considering these developments, we have a favorable scenario for the creation of a computational tool that integrates into one workflow all steps involved in molecular docking simulations. We had these goals in mind when we developed the program SAnDReS. This program allows the integration of all computational features related to modern docking studies into one workflow. SAnDReS not only carries out docking simulations but also evaluates several docking protocols allowing the selection of the best approach for a given protein system. SAnDReS is a free and open-source (GNU General Public License) computational environment for running docking simulations. Here, we describe the combination of SAnDReS and AutoDock4 for protein-ligand docking simulations. AutoDock4 is a free program that has been applied to over a thousand receptor-ligand docking simulations. The dataset described in this chapter is available for downloading at https://github.com/azevedolab/sandres.

Electrostatic Energy in Protein-Ligand Complexes.

Computational analysis of protein-ligand interactions is of pivotal importance for drug design. Assessment of ligand binding energy allows us to have a glimpse of the potential of a small organic molecule as a ligand to the binding site of a protein target. Considering scoring functions available in docking programs such as AutoDock4, AutoDock Vina, and Molegro Virtual Docker, we could say that they all rely on equations that sum each type of protein-ligand interactions to model the binding affinity. Most of the scoring functions consider electrostatic interactions involving the protein and the ligand. In this chapter, we present the main physics concepts necessary to understand electrostatics interactions relevant to molecular recognition of a ligand by the binding pocket of a protein target. Moreover, we analyze the electrostatic potential energy for an ensemble of structures to highlight the main features related to the importance of this interaction for binding affinity.

Van der Waals Potential in Protein Complexes.

Van der Waals forces are determinants of the formation of protein-ligand complexes. Physical models based on the Lennard-Jones potential can estimate van der Waals interactions with considerable accuracy and with a computational complexity that allows its application to molecular docking simulations and virtual screening of large databases of small organic molecules. Several empirical scoring functions used to evaluate protein-ligand interactions approximate van der Waals interactions with the Lennard-Jones potential. In this chapter, we present the main concepts necessary to understand van der Waals interactions relevant to molecular recognition of a ligand by the binding pocket of a protein target. We describe the Lennard-Jones potential and its application to calculate potential energy for an ensemble of structures to highlight the main features related to the importance of this interaction for binding affinity.

Hydrogen Bonds in Protein-Ligand Complexes.

Fast and reliable evaluation of the hydrogen bond potential energy has a significant impact in the drug design and development since it allows the assessment of large databases of organic molecules in virtual screening projects focused on a protein of interest. Semi-empirical force fields implemented in molecular docking programs make it possible the evaluation of protein-ligand binding affinity where the hydrogen bond potential is a common term used in the calculation. In this chapter, we describe the concepts behind the programs used to predict hydrogen bond potential energy employing semi-empirical force fields as the ones available in the programs AMBER, AutoDock4, TreeDock, and ReplicOpter. We described here the 12-10 potential and applied it to evaluate the binding affinity for an ensemble of crystallographic structures for which experimental data about binding affinity are available.

Molecular Dynamics Simulations with NAMD2.

X-ray diffraction crystallography is the primary technique to determine the three-dimensional structures of biomolecules. Although a robust method, X-ray crystallography is not able to access the dynamical behavior of macromolecules. To do so, we have to carry out molecular dynamics simulations taking as an initial system the three-dimensional structure obtained from experimental techniques or generated using homology modeling. In this chapter, we describe in detail a tutorial to carry out molecular dynamics simulations using the program NAMD2. We chose as a molecular system to simulate the structure of human cyclin-dependent kinase 2.

Docking with AutoDock4.

AutoDock is one of the most popular receptor-ligand docking simulation programs. It was first released in the early 1990s and is in continuous development and adapted to specific protein targets. AutoDock has been applied to a wide range of biological systems. It has been used not only for protein-ligand docking simulation but also for the prediction of binding affinity with good correlation with experimental binding affinity for several protein systems. The latest version makes use of a semi-empirical force field to evaluate protein-ligand binding affinity and for selecting the lowest energy pose in docking simulation. AutoDock4.2.6 has an arsenal of four search algorithms to carry out docking simulation including simulated annealing, genetic algorithm, and Lamarckian algorithm. In this chapter, we describe a tutorial about how to perform docking with AutoDock4. We focus our simulations on the protein target cyclin-dependent kinase 2.

Molegro Virtual Docker for Docking.

Molegro Virtual Docker is a protein-ligand docking simulation program that allows us to carry out docking simulations in a fully integrated computational package. MVD has been successfully applied to hundreds of different proteins, with docking performance similar to other docking programs such as AutoDock4 and AutoDock Vina. The program MVD has four search algorithms and four native scoring functions. Considering that we may have water molecules or not in the docking simulations, we have a total of 32 docking protocols. The integration of the programs SAnDReS ( https://github.com/azevedolab/sandres ) and MVD opens the possibility to carry out a detailed statistical analysis of docking results, which adds to the native capabilities of the program MVD. In this chapter, we describe a tutorial to carry out docking simulations with MVD and how to perform a statistical analysis of the docking results with the program SAnDReS. To illustrate the integration of both programs, we describe the redocking simulation focused the cyclin-dependent kinase 2 in complex with a competitive inhibitor.

Docking with GemDock.

GEMDOCK is a protein-ligand docking software that makes use of an elegant biologically inspired computational methodology based on the differential evolution algorithm. As any docking program, GEMDOCK has two major features to predict the binding of a small-molecule ligand to the binding site of a protein target: the search algorithm and the scoring function to evaluate the generated poses. The GEMDOCK scoring function uses a piecewise potential energy function integrated into the differential evolutionary algorithm. GEMDOCK has been applied to a wide range of protein systems with docking accuracy similar to other docking programs such as Molegro Virtual Docker, AutoDock4, and AutoDock Vina. In this chapter, we explain how to carry out protein-ligand docking simulations with GEMDOCK. We focus this tutorial on the protein target cyclin-dependent kinase 2.

Docking with SwissDock.

Protein-ligand docking simulation is central in drug design and development. Therefore, the development of web servers intended to docking simulations is of pivotal importance. SwissDock is a web server dedicated to carrying out protein-ligand docking simulation intuitively and elegantly. SwissDock is based on the protein-ligand docking program EADock DSS and has a simple and integrated interface. The SwissDock allows the user to upload structure files for a protein and a ligand, and returns the results by e-mail. To facilitate the upload of the protein and ligand files, we can prepare these input files using the program UCSF Chimera. In this chapter, we describe how to use UCSF Chimera and SwissDock to perform protein-ligand docking simulations. To illustrate the process, we describe the molecular docking of the competitive inhibitor roscovitine against the structure of human cyclin-dependent kinase 2.

Molecular Docking Simulations with ArgusLab.

Molecular docking is the major computational technique employed in the early stages of computer-aided drug discovery. The availability of free software to carry out docking simulations of protein-ligand systems has allowed for an increasing number of studies using this technique. Among the available free docking programs, we discuss the use of ArgusLab ( http://www.arguslab.com/arguslab.com/ArgusLab.html ) for protein-ligand docking simulation. This easy-to-use computational tool makes use of a genetic algorithm as a search algorithm and a fast scoring function that allows users with minimal experience in the simulations of protein-ligand simulations to carry out docking simulations. In this chapter, we present a detailed tutorial to perform docking simulations using ArgusLab.

Homology Modeling of Protein Targets with MODELLER.

Homology modeling is a computational approach to generate three-dimensional structures of protein targets when experimental data about similar proteins are available. Although experimental methods such as X-ray crystallography and nuclear magnetic resonance spectroscopy successfully solved the structures of nearly 150,000 macromolecules, there is still a gap in our structural knowledge. We can fulfill this gap with computational methodologies. Our goal in this chapter is to explain how to perform homology modeling of protein targets for drug development. We choose as a homology modeling tool the program MODELLER. To illustrate its use, we describe how to model the structure of human cyclin-dependent kinase 3 using MODELLER. We explain the modeling procedure of CDK3 apoenzyme and the structure of this enzyme in complex with roscovitine.