P159 Neun, an efficient and customizable open-source library for computational neural modeling and biohybrid circuit design.
Angel Lareo*1, Alicia Garrido-Peña1, Pablo Varona1, Francisco B. Rodriguez1
1Grupo de Neurocomputación Biológica, Departamento Ingeniería Informática, Universidad Autónoma de Madrid
*Email: angel.lareo@uam.es Introduction
Computational models are an effective and convenient tool for theoretically complementing the experimental results obtained from living systems and thus understanding the brain’s complex functions. Computational simulation of neural behavior has expanded the potential of modeling studies. There is a wide range of tools available in Neuroscience community for this purpose [1–6]. They have enhanced the ability of theoreticians to explain neural dynamics.Neunis a new highly customizable and fast running-time open-source framework designed for theoretical studies in single neurons, small circuits, and biohybrid circuit design [7–9]
Methods Neun(github.com/gnb-UAM/neun) is an object-oriented library with heavily templated C++ . This ensures high-level abstraction and encapsulation. Neun’s main components are: (i)ModelConcept, which provides the foundation for synapses and neuron models (e.g. Hodgkin-Huxley and Izhikevich paradigms). (ii) SystemWrapper, defines general elements such as parameters, variables, and numerical precision. (iii) Integrator, offer methods like Euler and Runge-Kutta for numerical integration. (iv)DifferentialNeuronWrapper, combines models and integrators for simulation.Neunalso uses a straightforward method for equation-to-code parsing to add new models and aims to provide compatibility with existing tools using a Python API. Results As a complement to existing tools and databases,Neunprovides built-in samples of well-known neuron and synapse models that can be easily adapted by the user for effective implementations. It can be used as a template for fast prototyping, since it offers boilerplate code for novel modelers. Users can then go from a black-box approach to the insides of the code. Nevertheless, the fact that the library is written in C++ makes it an attractive option for real-time applications (such as RTXI or embedded systems), as it demonstrates great single-threaded computing performance even without parallelization.Neunhas already been used in previous modeling studies [7–9] and has been tested for its use in real-time experiments. Discussion We presentNeun, an open-source library in C ++ for computational neural modeling and simulation as a user-friendly complement and alternative to existing tools. Among the numerous tools for neuron dynamics simulation, there is a tendency of increasing complexity in the code base which limits its accessibility, especially for beginners. We believeNeunhas a convenient compromise between usability and efficiency. This can be ideal for researchers in neuroscience who do not necessarily have a background in computer science but are willing to progressively learn, and also for experimentalist who want to build biohybrid circuits form interacting living and model neurons and synapses.
Acknowledgements This research was supported by grants PID2024-155923NB-I00, CPP2023-010818, PID2023-149669NB-I00, PID2021-122347NB-I00 (MCIN/AEI and ERDF – “A way of making Europe”). References [1]https://doi.org/10.1007/s10827-006-7949-5 [2]https://doi.org/10.3389/neuro.11.011.2008 [3]https://doi.org/10.1038/srep18854 [4]https://doi.org/10.7554/eLife.47314 [5]https://doi.org/10.1007/s10827-016-0623-7 [6]https://doi.org/10.1016/j.neuron.2019.05.019 [7]https://doi.org/10.3389/fninf.2022.912654 [8]https://doi.org/10.1007/978-3-031-34107-6_43 [9]https://doi.org/10.1117/1.NPh.11.2.024308
