Abstract:The computer-aided drug design field is constituted by a wide range of different approaches that can be roughly classified as either ligand- or receptor-based techniques. In order to design and develop new active ligands, ligand-based approaches use the information derived from the properties of ligands with known activity against the target of interest, whereas the receptor-based techniques use the information obtained from the knowledge of the three-dimensional structure of the target and from the analysis of ligand–receptor interactions. With regards to the receptor-based techniques, one of the key steps of almost all approaches is represented by the estimation of the binding strength between the ligand and the receptor. Indeed, a proper evaluation of this interaction can be profitably used to design new ligands characterized by a tighter binding with the target of interest. Unfortunately, the main limitation of docking calculation resides in the estimation of the ligand–receptor interaction, which is usually evaluated through simple scoring functions. In particular, the scoring functions are used for both qualitative and quantitative studies: they are used for ranking different binding poses of the same ligand, thus providing its most reliable binding orientation, and they can also be used for comparing from an energetical point of view the interaction of different ligands with the same receptor, thus ranking the ligands according to the relative binding affinities. Although the available scoring functions are able to suggest the most reliable binding orientation for a ligand, their ability to predict ligand-receptor binding affinity and correctly rank different ligands on this basis is often not satisfactory enough. Among the different approaches that can be applied to evaluate the ligand–receptor interaction, the accurate measurement of binding free energy plays a very important role and, in this field, the most rigorous theoretical method is represented by free-energy perturbation (FEP), which estimates the difference in free energy between two states by slowly changing one state to another through a number of nonphysical intermediate states, performing extended sampling at each intermediate state [<a href="#">1</a>]. Unfortunately, this method is extremely time consuming and therefore more approximate methods have been developed. In this context, the molecular mechanics Poisson–Boltzmann surface area (MM/PBSA) and molecular mechanics generalized Born surface area (MM/GBSA) developed by Kollman et al. [<a href="#">2</a>,<a href="#">3</a>] are among the most commonly used methods.

Drug interaction profiles classification: Sørensen–Dice coefficient based on MM/GBSA energies to explore focal adhesion kinase inhibitors

Using Interaction Fingerprints as a Strategy of Protein-Ligand Interaction Patterns Recognition and Post-Docking Analysis

Assessing the performance of docking, FEP, and MM/GBSA methods on a series of KLK6 inhibitors

Fast and accurate predictions of binding free energies using MM-PBSA and MM-GBSA

Fingerprinting Interactions between Proteins and Ligands for Facilitating Machine Learning in Drug Discovery

What is the current value of MM/PBSA and MM/GBSA methods in drug discovery?

Structural properties and interaction energies affecting drug design. An approach combining molecular simulations, statistics, interaction energies and neural networks

Drug—drug interaction through molecular structure similarity analysis

A molecular representation to identify isofunctional molecules

Comparative Evaluation of 11 Scoring Functions for Molecular Docking

A combination of spin diffusion methods for the determination of protein-ligand complex structural ensembles.

Combining Molecular Dynamic Information and an Aspherical-Atom Data Bank in the Evaluation of the Electrostatic Interaction Energy in Multimeric Protein-Ligand Complex: A Case Study for HIV-1 Protease

Combining the Fragment Molecular Orbital and GRID Approaches for the Prediction of Ligand-Metalloenzyme Binding Affinity: The Case Study of hCA II Inhibitors

Exploring the interaction between human focal adhesion kinase and inhibitors: a molecular dynamic simulation and free energy calculations

Combined Ensemble Docking and Machine Learning in Identification of Therapeutic Agents with Potential Inhibitory Effect on Human CES1

Enhance the Performance of Current Scoring Functions with the Aid of 3D Protein-Ligand Interaction Fingerprints.

SMPLIP-Score: predicting ligand binding affinity from simple and interpretable on-the-fly interaction fingerprint pattern descriptors

Comparison of Ligand-, Target Structure-, and Protein-Ligand Interaction Fingerprint-based Virtual Screening Methods

On Evaluating Molecular-Docking Methods for Pose Prediction and Enrichment Factors

ET‐score: Improving Protein‐ligand Binding Affinity Prediction Based on Distance‐weighted Interatomic Contact Features Using Extremely Randomized Trees Algorithm

Predicting Polypharmacology by Binding Site Similarity: From Kinases to the Protein Universe