Time: 15:00-16:30 on Thur.,Aug.17, 2023

Venue：E109, Biomedicine Hall

Speaker: Dr.Uygar Sumbul

Host: Dr.Xiaoxuan Jia

Title: Learning Neuronal Identity from Population Dynamics

Abstract：

Neurons can display highly variable dynamics. While such variability presumably supports the wide range of behaviors generated by the organism, their gene expressions are relatively stable in the adult brain. This suggests that neuronal activity is a combination of its time-invariant identity and the inputs the neuron receives from the rest of the circuit. Here, we propose an unsupervised deep learning-based method to assign time-invariant representations to individual neurons based on permutation-, and population size-invariant summary of population recordings. We fit dynamical models to neuronal activity to learn a representation by considering the activity of both the individual and the neighboring population. Our unsupervised approach and use of implicit representations enable robust inference against imperfections such as partial overlap of neurons across sessions, trial-to-trial variability, and limited availability of molecular (transcriptomic) labels for downstream supervised tasks. We demonstrate our method on a public multimodal dataset of mouse cortical neuronal activity and transcriptomic labels. 

Biography：

Uygar Sumbul is an Associate Investigator at the Allen Institute (USA). His research is broadly on computational neuroscience and machine learning. He obtained a PhD in Electrical Engineering and a PhD minor in Mathematics from Stanford University (USA) in 2009. Prior to joining the Allen Institute, he was a postdoctoral researcher at MIT (USA) with Sebastian Seung and at Columbia University (USA) with Liam Paninski. He obtained a BS in Electrical Engineering from Bilkent University (Turkey).