Benchmarking a Fast Proton Titration Scheme in Implicit Solvent for Biomolecular Simulations

Area of Science:

• Biophysics and computational biology
• Biomolecular modeling and simulation
• Protein chemistry

Background:

• pH is critical for biomolecular charge, influencing protein/RNA stability, enzyme activity, and interactions.
• Predicting protein structure and behavior is challenging due to complexity and limitations of existing computational methods.
• Technological applications in food, pharma, and biomaterials also rely on understanding pH-dependent processes.

Purpose of the Study:

• To benchmark a novel, fast proton titration scheme against experimental data and other theoretical methods.
• To evaluate the scheme's accuracy and robustness in predicting pKa values for diverse proteins.
• To assess the computational efficiency and ability to overcome sampling challenges in biomolecular simulations.

Main Methods:

• Developed a fast proton titration scheme based on the classical Tanford-Kirkwood model.
• Treated salt implicitly at the Debye-Hückel level to reduce computation time.
• Benchmarked the scheme against experimental pKa values for a set of representative proteins (HP36, BBL, HEWL, RNase, SNASE, ALAC, OMTKY3).

Main Results:

• Calculated pKa values showed average deviations of 0.4-0.9 pH units from experimental data.
• Maximum absolute and root-mean-square deviations were within acceptable ranges for theoretical models (1.0-5.2 and 0.5-1.2, respectively).
• The scheme outperformed the NULL model and, for specific proteins (BBL, ALAC, OMTKY3), provided more accurate predictions than other analyzed theoretical data.

Conclusions:

• The fast proton titration scheme is robust and accurately captures essential physics for biomolecular systems.
• Implicit salt treatment significantly reduces computation time and avoids sampling issues.
• This method offers a computationally efficient and reliable approach for predicting pH-dependent protein behavior.