| | An analytical guide map to identify hidden shared microcircuits
Tehran, Institute for Research in Fundamental Sciences (IPM):
Our brain has about 100 billion computing cells called neurons, each connected to thousands of other neurons. In the first glance, the neural network resembles the internet; each computer on the internet is connected to some other ones. However, all connections of a computer on the internet are clearly known to us whereas the connections of a neuron in the brain has, to a good extent, remained unknown.
The entangled and complex structure of neuronal network and the tiny width of axons and dendrites (which act like interconnecting wires among neurons) have a simple consequence: we can hardly identify map of neuronal connections, even for a volume of neuronal system as small as 1 micron cube. Particularly, when experimentalists deal with alive animals, it is very hard if not impossible to distinguish the neuronal microcircuit which functions among neurons. One way to find influential connections to a neuron is to record the activity of the neuron and all its inputs; this is not possible by existing methods. There are other techniques using large scale multi-electrode recordings or imaging methods, but they can only identify the connections among observed (recorded) neurons. However a large number of neurons remains unobserved.
Researchers of Institute for Research in Fundamental Sciences (IPM) jointly with Kyoto and Tehran Universities suggest a method to identify the hidden shared microcircuits from correlations among neurons’ activities, and solve the reverse problem. "We have learnt that similarity in the behavior of peoples, samples, or neurons, indicates their shared history. Here, behavior is the neuron’s activity, and history is the hidden shared inputs they receive.” illustrates S. Nader Rasuli of the School of Physics at IPM. "The aim of this research was to construct a theory that integrates the structure and activity of neural networks and use it as a tool to discover the network structure from the activity," says Hideaki Shimazaki of the Graduate School of Informatics, Kyoto University.
“ In this work, by using an analytical relation which traces the nonlinearity between input potential and output of neurons [1], we have linked different architectures with co-activities of neurons, then use it to identify the one influential microcircuit behind visual data.” says Safura Rashid Shomali of School of Cognitive Sciences at IPM.
The ability of brain for learning and information processing is attributed to the structure of the microcircuits which is shaped by neuronal connections and types. Therefore identifying the influential microcircuits changes our understanding about how brain works. “Our research proposes a new methodology to tackle this fundamental problem in brain science. The methodology is based on an interdisciplinary approach on the edge of brain science, physics and engineering.” Says Majid Nili Ahmadabadi of College of Engineering, University of Tehran.
The main result of this research is a guide map obtained from an accurate mathematical calculation. By calculating the interactions among neurons and comparing with the guide map, experimentalists can identify the structure and type (inhibitory/excitatory) of the hidden shared microcircuits in their data. “We use this guide map to reveal the architecture for monkey and mouse visual cortex. The results in both reveal that the motif of excitatory inputs shared between each pair of three neurons is behind the data. This is an example where theory unveils a mechanism that experiment cannot reach by its own” says coauthor Rasuli.
This method deals with co-activities among two and three neurons, and there still remains challenges for extending it to higher numbers. “It still remains a long way to solve the problem in general, but up to now, we hope that researchers use our guide map to determine the influential hidden microcircuits behind correlations among two or three neurons”, adds the corresponding author Shomali.
This research is published in (Nature) Communications Biology and is featured as the highlighted research of this week on the cover of the Journal [2]. https://www.nature.com/commsbio/
References:
1. Shomali, S.R., Ahmadabadi, M.N., Shimazaki, H., and Rasuli, S.N., How does transient signaling input affect the spike timing of postsynaptic neuron near the threshold regime: an analytical study. J Comput Neurosci 44, 147–171 (2018). https://doi.org/10.1007/s10827-017-0664-6
2. Shomali, S.R., Rasuli, S.N., Ahmadabadi, M.N., Shimazaki H., Uncovering hidden network architecture from spiking activities using an exact statistical input-output relation of neurons. Commun Biol 6, 169 (2023). https://doi.org/10.1038/s42003-023-04511-z
| |
|
