Evolving Simple Models of Diverse Intrinsic Dynamics in Hippocampal Neuron Types

Abstract

The diversity of intrinsic dynamics observed in neurons may enhance the computations implemented in the circuit by enriching network-level emergent properties such as synchronization and phase locking. Large-scale spiking network models of entire brain regions offer a platform to test theories of neural computation and cognitive function, providing useful insights on information processing in the nervous system. However, a systematic in-depth investigation requires network simulations to capture the biological intrinsic diversity of individual neurons at a sufficient level of accuracy. The computationally efficient Izhikevich model can reproduce a wide range of neuronal behaviors qualitatively. Previous studies using optimization techniques, however, were less successful in quantitatively matching experimentally recorded voltage traces. In this article, we present an automated pipeline based on evolutionary algorithms to quantitatively reproduce features of various classes of neuronal spike patterns using the Izhikevich model. Employing experimental data from Hippocampome.org, a comprehensive knowledgebase of neuron types in the rodent hippocampus, we demonstrate that our approach reliably fit Izhikevich models to nine distinct classes of experimentally recorded spike patterns, including delayed spiking, spiking with adaptation, stuttering, and bursting. Importantly, by leveraging the parameter-exploration capabilities of evolutionary algorithms, and by representing qualitative spike pattern class definitions in the error landscape, our approach creates several suitable models for each neuron type, exhibiting appropriate feature variabilities among neurons. Moreover, we demonstrate the flexibility of our methodology by creating multi-compartment Izhikevich models for each neuron type in addition to single-point versions. Although the results presented here focus on hippocampal neuron types, the same strategy is broadly applicable to any neural systems.

Simple models of quantitative firing phenotypes in hippocampal neurons: Comprehensive coverage of intrinsic diversity

“…On the other hand, highly abstract phenomenological models such as [6] specify only two equations, and they significantly reduce the computational cost of simulating large-scale networks [2,9,11]. However, the parameters that govern such models are not directly biologically interpretable and optimizing their parameters to reproduce quantitatively accurate intrinsic dynamics of neuron types can be difficult [12,13]. In current work, with a vision of creating a real-scale network model of the rodent hippocampus that nevertheless captures biological details at the mesoscopic level, we have created phenomenological models of 120 hippocampal neuron types and subtypes using their intrinsic dynamics identified experimentally.…”

Complex Dynamics in Simplified Neuronal Models: Reproducing Golgi Cell Electroresponsiveness

“…The model has only three state variables, the membrane potential and two currents, which can be associated to main biophysical subcellular mechanisms. Thanks to its mathematical structure, which is similar to GLIF and analytically solvable, E-GLIF can be optimized by traditional optimization algorithms (Pozzorini et al, 2015; Teeter et al, 2018) avoiding metaheuristic methods, like Genetic Algorithms, used for multi-compartment realistic neurons with high-dimensional parameter search space (Masoli et al, 2015) or non-linear LIF models (Venkadesh et al, 2018). E-GLIF can reproduce a rich variety of electroresponsive properties with a single set of optimal parameters: autorhythmicity, depolarization-induced excitation and post-inhibitory rebound bursting, specific input-output ( f-I stim ) relationships, spike-frequency adaptation, phase-reset, sub-threshold oscillations and resonance.…”

Optimization of Efficient Neuron Models With Realistic Firing Dynamics. The Case of the Cerebellar Granule Cell

“…Our optimization method is based on an EA that allows multiple parameter exploration to fit the experimentally recorded firing behavior (Jolivet et al, 2008 ; Hanuschkin et al, 2010 ; Barranca et al, 2013 ; Venkadesh et al, 2018 ). After the execution of the EA, it provides sets of parameters that minimize the FF, i.e., the function which associates each parameter set with a single value quantifying the goodness of such a neuron configuration.…”

“…Considering this, the adaptive exponential integrate-and-fire (AdEx) model (Brette and Gerstner, 2005 ) only includes two coupled differential equations that capture adaptation and resonance properties (Naud et al, 2008 ), while enabling large scale implementations of neuronal circuits. Although the AdEx model can be seen as a two-dimensional reduction of the spike initiation in HH models, the specific parameter values of the model configuration to match with electrophysiological measurements (Jolivet et al, 2008 ; Hanuschkin et al, 2010 ; Barranca et al, 2013 ; Venkadesh et al, 2018 ) cannot be experimentally determined as they require an automatic parameter tuning algorithm.…”