The EBRAINS Hodgkin-Huxley Neuron Builder: An online resource for building data-driven neuron models

Abstract

In the last decades, brain modeling has been established as a fundamental tool for understanding neural mechanisms and information processing in individual cells and circuits at different scales of observation. Building data-driven brain models requires the availability of experimental data and analysis tools as well as neural simulation environments and, often, large scale computing facilities. All these components are rarely found in a comprehensive framework and usually require ad hoc programming. To address this, we developed the EBRAINS Hodgkin-Huxley Neuron Builder (HHNB), a web resource for building single cell neural models via the extraction of activity features from electrophysiological traces, the optimization of the model parameters via a genetic algorithm executed on high performance computing facilities and the simulation of the optimized model in an interactive framework. Thanks to its inherent characteristics, the HHNB facilitates the data-driven model building workflow and its reproducibility, hence fostering a collaborative approach to brain modeling.

Different responses of mice and rats hippocampus CA1 pyramidal neurons to in vitro and in vivo-like inputs

“…For the full model optimization, we used the BluePyOpt tool ( Van Geit et al, 2016 ) embedded into a workflow available on the EBRAINS infrastructure ( Bologna et al, 2022 ), and in interactive use cases (see text footnote 1). BluePyOpt, is a multi-objective genetic algorithm that returns several suitable ensembles model parameters, in our case the peak ion channel conductances and passive properties, which best fit the experimental features.…”

Online interoperable resources for building hippocampal neuron models via the Hippocampus Hub

“…Data and models provided in the HH Build section can be selected and used to construct a data-driven single cell NEURON model, optimize its parameters and explore its behavior via in silico simulations, through the HHNB, 28 which is part of the ecosystem of tools and services available through the EBRAINS Research Infrastructure and has been thoroughly described in a previous work (Bologna et al, 2022). Briefly, the HHNB consists in a full stack web application that manages multi-user workflows that include three steps.…”

“…Then, the selection of a single neuron model, implemented in NEURON, from the MC 30 and/or the construction of a model via the upload of individual NEURON .mod and parameter files is done. Any .mod file can be uploaded by the users, but the content of these files must be mirrored in the parameters.json file (part of the BluePyOpt execution file ensemble and editable through the HHNB interface), where the neural mechanisms and the channel distributions are specified (see (Geit et al, 2016) and (Bologna et al, 2022) for further details). Finally, the optimization of the model parameters is performed via the genetic-algorithm-based Python library BluePyOpt (Geit et al, 2016); this process is run on HPC systems upon configuration of both the optimization algorithm parameters (e.g., number of generations) and the requested system resources (e.g., number of HPC nodes); (4) the simulation of the optimized model is launched in the BlueNeuronAsAService (BlueNaaS, https://ebrains-cls-interactive.github.io/online-use-cases.…”

“…In fact, we have built an ad hoc utility, the HHNB Service Account that allows to submit the model optimization jobs and manage individual quotas on behalf of the users, who are only requested to own EBRAINS credentials. In addition, those who possess an account on either the CSCS-DAINT or the NeuroScience Gateway (NSG), can use their own resources on these two systems for the optimization process -see (Bologna et al, 2022) for more details.…”

Can neuron modeling constrained by ultrafast imaging data extract the native function of ion channels?

“…Thus, realistic models of neurons and synapses can be assembled to simulate large scale networks (Markram et al, 2015 ) or they can be used to inspire artificial networks (Chavlis and Poirazi, 2021 ). More realistic models of neurons can be further developed by constraining the parameters of the model to more detailed experimental datasets using optimisation procedures (Van Geit et al, 2008 ; Iavarone et al, 2019 ; Bologna et al, 2022 ). Following this direction, neuronal computational tools can be eventually used to derive biophysical models of native ion channels from experimental observations (Cannon and D'Alessandro, 2006 ).…”