A global sampler of single particle tracking solutions for single molecule microscopy

Area of Science:

• Biophysics
• Computational Biology
• Molecular Dynamics

Background:

• Single particle tracking (SPT) often relies on model-fitting, limiting its ability to resolve complex cellular molecular motions.
• Heterogeneous molecular dynamics within cells present a significant challenge for traditional SPT evaluation methods.

Purpose of the Study:

• To develop a novel global spatiotemporal sampler for SPT solutions.
• To overcome limitations in resolving heterogeneous molecular motions using SPT.
• To enable accurate estimation of tracking error and provide alternative solutions.

Main Methods:

• Utilized a Metropolis-Hastings algorithm for a global spatiotemporal sampler.
• The algorithm assesses solution likelihood and presents alternative trajectories.
• It samples tracking parameters, eliminating the need for user tweaking.

Main Results:

• Demonstrated the algorithm's efficacy on both synthetic and single-molecule data.
• Generalized metrics for comparing SPT results to accommodate the new sampler.
• Showcased the propagation of tracking solution distributions into derived quantities, such as diffusion coefficients.

Conclusions:

• The developed sampler offers a more robust approach to analyzing particle trajectories in complex biological systems.
• It enhances the estimation of tracking errors and provides a distribution of possible solutions.
• Further discussion addresses the challenges in realizing advanced SPT samplers.