Compiling molecular ultrastructure into neural dynamics
Abstract
High-resolution brain imaging can now capture not just synapse locations but their molecular composition, with the cost of such mapping falling exponentially. Yet such ultrastructural data has so far told us little about local neuronal physiology - specifically, the parameters (e.g., synaptic efficacies, local conductances) that govern neural dynamics. We propose to translate molecularly annotated ultrastructure into physiology, introducing the concept of an ultrastructure-to-dynamics compiler: a learned mapping from molecularly annotated ultrastructure to simulator-ready, uncertainty-aware physiological parameters. The requirement is paired training data, with jointly acquired ultrastructure from imaging, and dynamical responses to perturbations from physiological experiments. With this data we can train models that predict local physiology directly from structure. Such a compiler would support biophysical simulations by turning anatomical maps into models of circuit dynamics, shifting structure-to-function from a descriptive program to a predictive one and opening routes to understanding neural computation and forecasting intervention effects.
Cite
@article{arxiv.2603.25713,
title = {Compiling molecular ultrastructure into neural dynamics},
author = {Konrad P. Kording and Anton Arkhipov and Davy Deng and Sean Escola and Seth G. N. Grant and Gal Haspel and Michał Januszewski and Narayanan Kasthuri and Nina Khera and Richie E. Kohman and Grace Lindsay and Jeantine Lunshof and Adam Marblestone and David A. Markowitz and Jordan Matelsky and Brett Mensh and Patrick Mineault and Andrew Payne and Joanne Peng and Xaq Pitkow and Philip Shiu and Gregor Schuhknecht and Sven Truckenbrodt and Joshua T. Vogelstein and Edward S. Boyden},
journal= {arXiv preprint arXiv:2603.25713},
year = {2026}
}