A clustering-based biased Monte Carlo approach to protein titration curve prediction

TL;DR

This paper introduces a clustering-based biased Monte Carlo method to efficiently predict protein titration curves by improving sampling in systems with strong energetic interactions.

Contribution

It presents a novel biased MCMC approach leveraging cluster structures to enhance sampling efficiency in protein titration curve prediction.

Findings

Biased MCMC outperforms regular MCMC in efficiency.

Method validated on synthetic and real protein systems.

Improved accuracy in estimating protonation ensemble averages.

Abstract

In this work, we developed an efficient approach to compute ensemble averages in systems with pairwise-additive energetic interactions between the entities. Methods involving full enumeration of the configuration space result in exponential complexity. Sampling methods such as Markov Chain Monte Carlo (MCMC) algorithms have been proposed to tackle the exponential complexity of these problems; however, in certain scenarios where significant energetic coupling exists between the entities, the efficiency of the such algorithms can be diminished. We used a strategy to improve the efficiency of MCMC by taking advantage of the cluster structure in the interaction energy matrix to bias the sampling. We pursued two different schemes for the biased MCMC runs and show that they are valid MCMC schemes. We used both synthesized and real-world systems to show the improved performance of our biased MCMC methods when compared to the regular MCMC method. In particular, we applied these algorithms to the problem of estimating protonation ensemble averages and titration curves of residues in a protein.