A Pipeline for the Creation of Biophysically Realistic Multicompartment Models of Drosophila MDN

Wednesday, June 22, 2022

2:00 PM-4:00 PM

BIOMED Master's Thesis Defense

Title:
A Pipeline for the Creation of Biophysically Realistic Multicompartment Models of Drosophila Melanogaster Descending Neurons (MDN)

Speaker:
Alexander N. Vasserman, Master’s Candidate
School of Biomedical Engineering, Science and Health Systems
Drexel University

Advisors:
Jessica Ausborn, PhD
Assistant Professor
Department of Neurobiology and Anatomy
College of Medicine
Drexel University

Catherine von Reyn, PhD
Assistant Professor
School of Biomedical Engineering, Science and Health Systems
Drexel University

Vikas Bhandawat, PhD
Associate Professor
School of Biomedical Engineering, Science and Health Systems
Drexel University
 

Details:
Sensorimotor transformations are neural computations that convert high-dimensional sensory input into motor commands. Understanding how this conversion occurs will give insight into how organisms select appropriate behaviors in response to sensory stimuli. In the Drosophila melanogaster escape circuit, sensorimotor transformations occur in descending neurons (DNs) that receive inputs from feature encoding visual projection neurons (VPNs) and convey motor commands onto downstream motor circuits. This pivotal role, together with a rich suite of genetic and electrophysiological tools, makes DNs ideal models for sensorimotor transformations. We recently identified a bilateral pair of DNs (DNp03) that exhibits persistent activity (PA) when presented with visual inputs that mimic the looming shadow of a predator.

The role of PA in transforming looming evoked sensory inputs into coordinated motor outputs in DNp03, however, is unknown. Detailed biophysical multicompartment models can be powerful tools to study how electrophysiological characteristics of a neuron such as PA shape its computations. As a first step, we develop a pipeline to create biophysically accurate computational models of Drosophila neurons and validate it by developing a model of DNp03. Utilizing this model we investigate the organization of sensory inputs DNp03 receives from VPNs and analyze their passive propagation. To guide implementation of active properties we use selective fluorescent labeling of sodium channels in DNp03 and discover its putative spike-initiation zone. The model can now be used to investigate the role of individual ion channels in sensorimotor integration and the modeling pipeline will enable the fast implementation of model neurons in the future.

Contact Information

Natalia Broz
njb33@drexel.edu