BRAINCELL: A modelling platform for stochastic nanoscale organisation and intra- and intercellular signalling in neurons and glia

Abstract

Biophysical cell models have been key to our understanding of information processing in neurons and their networks. However, as brain computation emerges from the complex interaction among neuronal and non-neuronal cells and their microenvironment, critical limitations have persisted in existing modelling approaches. Firstly, most brain cells exhibit a rich repertoire of nanoscopic structures, such as dendritic spines in neurons or nanoscopic protrusions in astrocytes, which have been difficult to incorporate into whole-cell models because of their numbers and complexity. To address this, BRAINCELL introduces a stochastic framework for generating populations of nanoscale morphological and physiological features grounded in empirical statistics. Secondly, physiological activity of brain cells critically depends on their dynamic interactions with the extracellular environment, which has traditionally been treated as static. BRAINCELL creates an interactive extracellular environment that tracks concentration dynamics for ions and signalling molecules inside and outside cells. With key BRAINCELL algorithms tested and validated in our previous experimental studies, we show that these advances enable realistic modelling of cell-cell interactions mediated by the extracellular space, such as between neurons and astrocyte or between axons and myelin, opening new avenues for testing the biophysical basis of empirical observations. By integrating morphological and functional stochasticity with dynamics extracellular interactions, BRAINCELL enables task-specific predictions that often diverge from those produced by conventional models. BRAINCELL is freely available at www.neuroalgebra.net.