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

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.

Identification of allosteric inhibitor against AKT1 through structure-based virtual screening

AKT (serine/threonine protein kinase) is a potential therapeutic target for many types of cancer as it plays a vital role in cancer progression. Many AKT inhibitors are already in practice under single and combinatorial therapy. However, most of these inhibitors are orthosteric / pan-AKT that are non-selective and non-specific to AKT kinase and their isoforms. Hence, researchers are searching for novel allosteric inhibitors that bind in the interface between pH and kinase domain. In this study, we performed structure-based virtual screening from the afroDB (a diverse natural compounds library) to find the potential inhibitor targeting the AKT1. These compounds were filtered through Lipinski, ADMET properties, combined with a molecular docking approach to obtain the 8 best compounds. Then we performed molecular dynamics simulation for apoprotein, AKT1 with 8 complexes, and AKT1 with the positive control (Miransertib). Molecular docking and simulation analysis revealed that Bianthracene III (hit 1), 10-acetonyl Knipholonecyclooxanthrone (hit 2), Abyssinoflavanone VII (hit 5) and 8-c-p-hydroxybenzyldiosmetin (hit 6) had a better binding affinity, stability, and compactness than the reference compound. Notably, hit 1, hit 2 and hit 5 had molecular features required for allosteric inhibition.