DATE2022.11.01 #Press Releases
Detecting single synapses from activity data of dense synaptic populations using
statistical mechanics methods
Disclaimer: machine translated by DeepL which may contain errors.
The University of Electro-Communications
Graduate School of Frontier Sciences, The University of Tokyo
Graduate School of Science, The University of Tokyo
Summary of Presentations
A research group led by Associate Professor Hiroshi Kosaka of the Graduate School of Information Science and Technology, the University of Electro-Communications, graduate student Kazushi Fukumasu of the Department of Physics, Graduate School of Science of the University of Tokyo, and Professor Satonao Nose of the Department of Complexity Science and Engineering of Frontier Sciences, the Graduate School of Frontier Sciences, has successfully extracted synaptic activity from calcium imaging data of complex CNS systems by image analysis. They have succeeded in extracting synaptic-level activity from calcium imaging data of the complex central nervous system.
Neural circuits consist of a large number of neurons, which process a variety of information, including sensory reception, motor control, memory, and thought (Figure 1). Such information processing is accomplished by individual neurons sending signals to other neurons via microscopic structures called synapses. Therefore, to elucidate the mechanism of information processing, the activity of countless synapses must be measured simultaneously. Although simultaneous activity measurement of the cell body, the main body of a neuron, has been conducted in the past, a method that enables large-scale measurement of the synaptic population itself has not been established.
In this study, we developed a method for detecting synaptic population activity using the motor control circuit of Drosophila larvae, whose neural circuits have been studied at the cellular level. First, using a calcium imaging probe molecule that can capture the neural activity of individual synapses as changes in fluorescence intensity, we captured fluorescent images of neural activity during the generation of motor control signals in the central nervous system (Figure 2 left). From these fluorescent time series images, which show the complex activity of thousands of synapses, we developed an algorithm (PQ-clustering) to isolate individual synapses on the image and extract their temporal activity patterns by using powerful mathematical methods of graph theory and statistical physics.
The efficient method developed in this study for extracting the activity of large synaptic populations is expected to further elucidate the information processing mechanisms in neural circuits.
The results of this research study were published in the international scientific journal Neural Networks on October 6.
For more information, please visit the website of the University of Electro-Communications.