A universal workflow for creation, validation, and generalization of detailed neuronal models

• DOI: 10.1016/j.patter.2023.100855
• Publication Date: 2023-10-04T14:28:40Z
• Authors: Maria Reva; Christian Rössert; Alexis Arnaudon; Tanguy Damart; Darshan Mandge; Anıl Tuncel; Srikanth Ramaswamy; Henry Markram; Werner Van Geit
• Keywords: Computational model; Biological neuron model

Detailed single-neuron modeling is widely used to study neuronal functions. While cellular and functional diversity across the mammalian cortex vast, most of available computational tools focus on a limited set specific features characteristic single neuron. Here, we present generalized automated workflow for creation robust electrical models illustrate its performance by building cell rat somatosensory cortex. Each model based 3D morphological reconstruction ionic mechanisms. We use an evolutionary algorithm optimize parameters match electrophysiological extracted from experimental data. Then validate optimized against additional stimuli assess their generalizability population similar morphologies. Compared state-of-the-art canonical models, our show 5-fold improved generalizability. This versatile approach can be build any type.