BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Computational Neuroscience Journal Club - Rui Xia and David Lui
DTSTART:20230606T130000Z
DTEND:20230606T150000Z
UID:TALK202243@talks.cam.ac.uk
CONTACT:Luke Johnston
DESCRIPTION:Please join us for our fortnightly Computational Neuroscience 
 journal club on Tuesday 6th June at 2pm UK time in the CBL seminar room\, 
 or online on zoom. The title is ‘Switching models to uncover hidden beha
 vioural and neural states’\, presented by Rui Xia and David Lui.\n\nZoom
  information: https://eng-cam.zoom.us/j/84204498431?pwd=Um1oU284b1YxWThObG
 w4ZU9XZitWdz09 Meeting ID: 842 0449 8431 Passcode: 684140\n\nSummary:\n\nA
 dvances in modern recording technologies have enable large-scale measureme
 nts of neural activity in orders of magnitude more than we could only a fe
 w years ago. These datasets offer unprecedented opportunities to study how
  neural circuits function\, process sensory information and generate behav
 iour. To gain better insight to these complex and heterogeneous time serie
 s data with nonlinear dynamics\, one approach is decomposing the data into
  segments that each can be explained by simple\, linear dynamics. Linderma
 n et al [1] proposed recurrent switching linear dynamical systems which ca
 n automatically divide latent space of population neural activities into d
 iscrete\, behaviourally significant states. We will present two papers app
 lying this flexible and interpretable model to recordings of head ganglia 
 neurons in nematode C. elegans [2]\, and neuronal subpopulations within MP
 OA (medial preoptic area) and VMHvl (ventromedial hypothalamus) in mice du
 ring innate social behaviours [3]. In the experiments of C. elegans\, the 
 framework reveals states that closely match manual labels of different beh
 aviours\, such as forward crawling\, reversals and turns. For hypothalamus
  recordings in mice\, an approximate line attractor is uncovered\, progres
 sion along which is discovered to be correlated with an escalation of agon
 istic behaviour.\n\nReferences:\n\n[1] Linderman\, Scott\, et al. “Bayes
 ian learning and inference in recurrent switching linear dynamical systems
 .” Artificial Intelligence and Statistics. PMLR\, 2017.\n\n[2] Linderman
 \, Scott\, et al. “Hierarchical recurrent state space models reveal disc
 rete and continuous dynamics of neural activity in C. elegans.” BioRxiv 
 (2019): 621540.\n\n[3] Nair\, Aditya\, et al. “An approximate line attra
 ctor in the hypothalamus encodes an aggressive state.” Cell 186.1 (2023)
 : 178-193.
LOCATION:In Person (CBL Seminar Room) and Online on Zoom
END:VEVENT
END:VCALENDAR