Abstract
SummaryInformation processing in the brain depends on the integration of synaptic input distributed throughout neuronal dendrites. Dendritic integration is a hierarchical process, proposed to be equivalent to integration by a multilayer network, potentially endowing single neurons with substantial computational power. However, whether neurons can learn to harness dendritic properties to realize this potential is unknown. Here, we develop a learning rule from dendritic cable theory and use it to investigate the processing capacity of a detailed pyramidal neuron model. We show that computations using spatial or temporal features of synaptic input patterns can be learned, and even synergistically combined, to solve a canonical nonlinear feature-binding problem. The voltage dependence of the learning rule drives coactive synapses to engage dendritic nonlinearities, whereas spike-timing dependence shapes the time course of subthreshold potentials. Dendritic input-output relationships can therefore be flexibly tuned through synaptic plasticity, allowing optimal implementation of nonlinear functions by single neurons.
Highlights
An essential role of each neuron in a circuit is to transform a barrage of synaptic input into a meaningful stream of action potential output
It has been proposed that the combination of dendritic morphology and local NMDA receptor-dependent nonlinearities form a hierarchical processing structure with substantial computational power (Mel, 1992b; Archie and Mel, 2000; Poirazi and Mel, 2001)
Our approach is general, we focus on understanding how dendritic morphology and NMDA receptor-dependent excitability can be recruited without requiring structured connectivity
Summary
An essential role of each neuron in a circuit is to transform a barrage of synaptic input into a meaningful stream of action potential output. In the first stage of processing, synaptic input is integrated nonlinearly within individual dendrites, followed by a subsequent stage in which current flowing from dendrites is integrated at the soma, elegantly summarized as an equivalence of single neurons to multilayer neural networks (Poirazi et al, 2003a, 2003b; Jadi et al, 2014; Ujfalussy et al, 2018; Beniaguev et al, 2021; Jones and Kording, 2021a) This influential theory is incomplete, as it is yet to be comprehensively determined how or whether computations that capitalize on dendritic physiology can be learned
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
