Data-driven classification of individual cells by their non-Markovian motion

Area of Science:

• Biophysics
• Cell Biology
• Microbiology

Background:

• Understanding organismal motion is crucial for various biological and medical applications.
• The generalized Langevin equation (GLE) provides a framework for modeling active or passive particle and organismal motion, incorporating memory effects (non-Markovian dynamics).
• Distinguishing different motile behaviors at the single-cell level is a significant challenge.

Purpose of the Study:

• To introduce a novel method for differentiating organisms based exclusively on their motion trajectories.
• To apply this method to distinguish between two distinct swimming modes of the unicellular microalgae Chlamydomonas reinhardtii.
• To demonstrate the broad applicability of motion-based classification in biological and medical contexts.

Main Methods:

• Utilizing the generalized Langevin equation (GLE) to model organismal motion.
• Extracting all relevant GLE parameters from individual cell trajectories.
• Employing unbiased cluster analysis to categorize cells into distinct groups based on their motion parameters.

Main Results:

• The GLE-based method successfully differentiated between two distinct swimming modes of Chlamydomonas reinhardtii.
• Parameter extraction and cluster analysis provided a robust classification of cell behaviors.
• Control experiments validated the accuracy of the GLE-based assignment into swimming modes.

Conclusions:

• Motion trajectory analysis using the generalized Langevin equation offers a powerful tool for organism classification.
• This method provides an unbiased approach to distinguish between different motile behaviors.
• The technique has significant potential for applications in cell sorting, diagnostics, and fundamental biological research.