NEST framework brings dendritic dynamics into large-scale neural simulations
Published in Brain/Neurology, Journals, Numerical Simulations, Supercomputing/Calculations, Tools.
Published October 28, 2026.
A new framework reported October 27 in Communications Biology makes it easier to include dendritic processes in large neural-network simulations. Wybo and colleagues extend NEST, a simulator commonly used for networks of simplified spiking neurons.
The framework lets researchers describe customizable subcellular dynamics in NESTML and generates efficient C++ implementations. Neurons can then contain multiple compartments, supporting models that combine network scale with selected details of dendritic signaling.
Benchmarks demonstrated efficient parallel execution. As a proof of concept, a simplified network used apical calcium dynamics to associate contextual input arriving at distal dendrites with sensory input arriving nearer the cell body in a single-shot learning task.
This provides a computational tool for testing how cellular mechanisms influence network behavior. The learning demonstration depends on the chosen equations and assumptions; it does not establish a complete biological account of memory or recreate a human brain. Reported speed advantages also depend on the tested models, hardware and simulation settings.
Source: Wybo et al., Communications Biology 9, 1411, October 27, 2026. DOI: 10.1038/s42003-026-11157-0.