Three-Dimensional-QSAR and Relative Binding Affinity Estimation of Focal Adhesion Kinase Inhibitors

Area of Science:

• Computational chemistry
• Drug discovery
• Molecular modeling

Background:

• Precise binding affinity predictions are crucial for structure-based drug discovery (SBDD).
• Focal adhesion kinase (FAK) is a key target in oncology due to its overexpression in various cancers.
• FAK inhibition represents a promising therapeutic strategy for cancer treatment.

Purpose of the Study:

• To computationally model and predict the binding affinity of small molecule inhibitors targeting FAK.
• To explore the synergy between machine learning (ML) and physics-based approaches for optimizing FAK inhibitors.

Main Methods:

• Employed three-dimensional quantitative structure-activity relationship (3D-QSAR) methods (CoMFA, CoMSIA) utilizing supervised ML.
• Utilized molecular dynamics (MD) simulations and MM-PB/GBSA for analyzing binding interactions.
• Applied alchemical free energy perturbation (FEP) simulations to estimate relative binding free energies.

Main Results:

• 3D-QSAR studies demonstrated reasonable statistical accuracy in correlating physicochemical properties with inhibitory activities.
• MD simulations provided insights into residue-specific binding interactions within the FAK active site.
• Computed relative binding free energies showed acceptable agreement with experimental data.

Conclusions:

• Hybrid approaches combining ML (3D-QSAR) and physics-based simulations (MD, FEP) are effective for rational drug design.
• These integrated computational strategies can accelerate the optimization of lead compounds targeting FAK.
• The study highlights the potential of synergistic computational methods in structure-based drug discovery for oncology.