Abstract:The characterization of network and biophysical properties from neural spiking activity is an important goal in neuroscience. A framework that provides unbiased inference on causal synaptic interaction and single neural properties has been missing. Here we applied the stochastic dynamics extension of Maximum Entropy -- the Maximum Caliber Principle -- to infer the transition rates of network states. Effective synaptic coupling strength and neuronal response functions for various network motifs can then be computed. The inferred minimal model also enables leading-order reconstruction of inter-spike interval distribution. Our method is tested with numerical simulated spiking networks and applied to data from salamander retina.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: to infer network and biophysical properties from the spiking activity of neurons (i.e., spike trains), especially in the field of neuroscience, to provide an unbiased inference framework to determine causal synaptic interactions and the properties of individual neurons. Specifically, the authors applied the dynamic extension of the maximum entropy principle - the Maximum Caliber Principle (Max Cal) - to infer the transition rates of network states and calculate the effective synaptic coupling strength and neuron response functions from them. ### Core problems of the paper 1. **Inferring synaptic coupling strength and neuron response functions** - The authors aim to infer the effective synaptic coupling strength and neuron response functions through the Maximum Caliber Principle. - These inferences can help understand the information - transfer mechanisms in neural networks and the interactions between neurons. 2. **Constructing the minimal dynamic model** - The minimal dynamic model constructed through the Maximum Caliber Principle can capture the main features of the spike - interval distribution. - This model can be used to reconstruct and predict the behavior of neural networks. 3. **Verifying the effectiveness of the method** - The authors tested their method using numerical simulation of spiking network data and applied it to the data of zebrafish retina to identify different network motifs. ### Formula explanations - **Jump rate \( R_{x,y} \)**: It represents the rate of jumping from network state \( x \) to state \( y \), and is defined as: \[ R_{x,y}=\lim_{T\rightarrow\infty}\frac{C_{x,y}}{\tau_x} \] where \( C_{x,y} \) is the number of transitions from state \( x \) to state \( y \), and \( \tau_x \) is the total occupancy time of state \( x \). - **Effective synaptic coupling strength \( w_{j,i} \)**: It is defined by comparing the ratios of jump rates in different states, for example: \[ w_{j,i}=\ln\left(\frac{R_{(010),(110)}}{R_{(000),(100)}}\right) \] where \( w_{j,i} \) represents the synaptic coupling strength of neuron \( j \) to neuron \( i \). - **Effective neuron response function \( \phi(I_i) \)**: It is constructed based on the network - state jump rate \( R_{\xi,\eta} \) and represents the firing rate of neuron \( i \) under different input currents \( I_i \): \[ f_i = \phi(0),\quad f_i^{w_{j,i}}=\phi(w_{j,i}),\quad f_i^{w_{jk,i}}=\phi(w_{j,i}+w_{k,i}) \] ### Results and applications - **Verification and application**: The authors verified the effectiveness of the method by simulating different types of neural network motifs (such as common - input, chain, and cyclic motifs) and demonstrated its potential for application in processing actual biological data (such as zebrafish retina data). In summary, this paper proposes a method based on the Maximum Caliber Principle for inferring network and biophysical properties from neuronal spiking activities, especially synaptic coupling strength and neuron response functions. This method not only helps to understand the information - transfer mechanisms of neural networks but can also be applied to the analysis of actual biological data.

Maximum Caliber Infers Effective Coupling and Response from Spiking Networks

Exact computation of the Maximum Entropy Potential of spiking neural networks models

Future societal issues in industrial biotechnology

A causal inference approach of monosynapses from spike trains

Efficient Inference on a Network of Spiking Neurons using Deep Learning

Biologically plausible solutions for spiking networks with efficient coding

Maximum Entropy Principle Analysis in Network Systems with Short-time Recordings.

Exact minimax entropy models of large-scale neuronal activity

Inference of network connectivity from temporally binned spike trains

Searching for collective behavior in a network of real neurons

Maximum entropy models reveal the excitatory and inhibitory correlation structures in cortical neuronal activity

Scaling of unitary synaptic strength in the context of network dynamics

Excitation–Inhibition Balance, Neural Criticality, and Activities in Neuronal Circuits

Reconstruction of recurrent synaptic connectivity of thousands of neurons from simulated spiking activity

Equivalence of Additive and Multiplicative Coupling in Spiking Neural Networks

Common-input models for multiple neural spike-train data

Stimulus-dependent maximum entropy models of neural population codes

Spontaneous and stimulus-induced coherent states of critically balanced neuronal networks

TARGETING LOWER LIMB FUNCTION IN CHILDREN WITH CEREBRAL PALSY : A SYSTEMATIC REVIEW USING THE INTERNATIONAL CLASSIFICATION OF FUNCTIONING , DISABILITY AND HEALTH AS A FRAMEWORK

Approximating Nonlinear Functions With Latent Boundaries in Low-Rank Excitatory-Inhibitory Spiking Networks

Response of electrically coupled spiking neurons: A cellular automaton approach