A time-efficient computational binding affinity estimation protocol with utilization of limited experimental data: A case study for adenosine receptor

Abstract

Estimating binding affinity is a crucial step in the drug discovery process. In computer-aided drug design, this challenge can be divided into two main tasks: finding the correct binding pose and estimating the binding free energy. In this study, we propose a new binding affinity estimation protocol that utilizes molecular docking with limited experimental data and estimates binding affinity using molecular dynamics simulation. A custom scoring function was employed during docking to identify an improved initial binding pose, and the linear interaction energy method with an optimized coefficient was used for binding affinity estimation. The protocol was validated with an external data set and applied to modafinil and its derivatives to rank their binding affinities to adenosine A2A receptors (ADORA2A) as a case study. This approach could be both time-efficient and valuable for computational drug discovery, particularly when experimental data is limited.