Thermodynamic potential calculations via classical molecular dynamics: a virtual screening strategy for identifying mGluR7/8 and 5-HT1A agonists for Autism Spectrum Disorder

Abstract

The use of computational resources for thermodynamic calculations has been a revolution in chemical sciences and plays an indispensable role in the study of macromolecular biological systems, such as protein/enzyme receptors and small ligands. The implementation of rapid calculations, based on implicit solvent models and approximations of statistical ensembles through classical molecular dynamics, has gained prominence due to its numerically feasible results and drastically reduced computational cost compared to standard methods such as thermodynamic integration and perturbative calculations. The central application proposed in this work lies in virtual screening protocols, which require fast and efficient computational calculations to test various pharmacologically viable ligands against their respective biological targets. The methodology employed involves the search for ligands that are structurally and/or pharmacophorically similar to reference compounds in publicly accessible molecular libraries, statistically validated molecular docking calculations with discriminatory power for biological activity, construction of QSAR models based on machine learning, predictions of biological and physicochemical ADMET (absorption, distribution, metabolism, excretion, and toxicity) properties, classical molecular dynamics simulations, and free energy calculations using QM/MM-PB/GBSA and free energy landscapes. In this work, two scientific manuscripts were proposed using these methodologies, with application to the discovery of agonists for mGluR7/8 and 5-HT1A, proteins highly correlated with the pathogenesis of autism spectrum disorder (ASD) in humans – a neuropsychiatric condition affecting 1 in 100 children worldwide, characterized by heterogeneous symptoms such as restricted social behavior, repetitive gestures, speech and cognitive difficulties, and severe comorbidities like depression, anxiety, and suicidal ideation. The application of fast and effective computational methods is of utmost importance, given the urgent nature of this issue and the lack of safe chemotherapeutic treatments. As a result of these studies, we identified three dual agonists of mGluR7/8 with established computational activity and three selective 5-HT1A agonists, two of which are already in advanced stages of clinical testing.