Fast Bayesian inference of optical trap stiffness and particle diffusion.

Area of Science:

• Physics
• Physical Chemistry
• Statistical Mechanics

Background:

• Stochastic processes, like Brownian motion, are fundamental in understanding particle dynamics.
• The Ornstein-Uhlenbeck process accurately models overdamped Brownian motion in optical traps.
• Experimental observation of particle movement provides data for parameter estimation.

Purpose of the Study:

• To present novel Bayesian inference methods for parameter estimation in optical trap experiments.
• To accurately determine trap stiffness and particle diffusion coefficients.
• To offer a computationally efficient alternative to traditional methods.

Main Methods:

• Utilizing Bayesian inference with exact likelihoods and sufficient statistics.
• Deriving simple expressions for maximum a posteriori (MAP) estimates.
• Avoiding computationally intensive Monte Carlo sampling techniques.

Main Results:

• Developed fast and accurate Bayesian methods for parameter estimation.
• Achieved simple, closed-form expressions for maximum a posteriori estimates.
• Demonstrated superior performance compared to non-Bayesian fitting methods on experimental data.

Conclusions:

• The presented Bayesian approach offers a principled and efficient framework for analyzing optical trap data.
• These methods provide accurate estimations of trap stiffness and diffusion coefficients.
• The approach significantly advances the analysis of stochastic processes in experimental physics.