BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Data-driven classification of single cells by their non-markovian 
 motion - Anton Klimek (Freie Universität Berlin)
DTSTART:20231110T093000Z
DTEND:20231110T100000Z
UID:TALK203971@talks.cam.ac.uk
DESCRIPTION:The motion of organisms is well known to be a non-equilibrium 
 process\, which exhibits drastically different features for different type
 s of organisms. From an evolutionary perspective one can interpret organis
 ms&#39\; patterns of motion as adaptions to search efficiently for resourc
 es [Klimek & Netz EPL 2022] and it is known that cell motion can exhibit n
 on-Markovian motion\, which be described by the generalized Langevin equat
 ion (GLE) [Mitterwallner et al. PRE 2020].\nHere\, we present a method to 
 differentiate cells solely by their non-Markovian trajectories based on th
 e GLE in a non-equilibrium framework and apply it to distinguish two diffe
 rent swimming types of strongly confined microalgae Chlamydomonas reinhard
 tii cells with an accuracy of 100%.\nThe model we use is suggested by the 
 data and succeeds to describe the motion on the single cell level. By a si
 mple fit we can extract model parameters for individual cells and subseque
 ntly perform an unbiased cluster analysis to determine the number of diffe
 rent cell types in the population and obtain an assignment of every cell t
 o one of the types. Additionally\, the model suggested by the data include
 s information on the underlying processes leading to the observed patterns
  of motion\, which in the case of our Chlamydomonas reinhardtii data hints
  towards a harmonic coupling inside of the cell.\nAs it still remains a ch
 allenge to classify cells on the single cell level\, the presented method 
 to distinguish cells with as little information as their trajectories migh
 t have important implications in biology and medicine.
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR