Comparative Assessment of Free Energy Computational Methods for Revealing the Interactions Driving PARP1 Selective...

Area of Science:

• Computational chemistry
• Drug discovery
• Structural biology

Background:

• Accurate prediction of inhibitor selectivity across protein paralogues is crucial for drug discovery.
• PARP enzyme inhibitors are vital for treating various cancers, including ovarian, breast, and prostate tumors.

Purpose of the Study:

• To comparatively assess three computational methods (MM/PBSA, ABFE, and US) for predicting PARP1 versus PARP2 inhibitor selectivity.
• To evaluate the ability of these methods to recapitulate experimental binding affinities for eight clinically relevant PARP inhibitors.

Main Methods:

• Molecular Mechanics/Poisson-Boltzmann Surface Area (MM/PBSA) calculations.
• Absolute Binding Free Energy (ABFE) calculations.
• Umbrella Sampling (US) calculations with atomistic models and explicit solvent.

Main Results:

• MM/PBSA provides rapid insights but is sensitive to conformational poses, challenging for subtle energetic differences.
• ABFE and US calculations show improved agreement with experimental binding affinities.
• ABFE demonstrated the strongest quantitative correlation with experimental binding free energy differences, accurately reproducing selectivity trends.

Conclusions:

• ABFE and US calculations offer reliable predictions of inhibitor selectivity, outperforming MM/PBSA.
• Structural contact analysis provides mechanistic insights into ligand selectivity and identifies key stabilizing residues.
• This study informs the design of more potent and specific inhibitors for precision oncology, particularly for homologous recombination-deficient cancers.