Single neuron dynamics and computation

•LNP models provide a precise description of the static and dynamic responses of neurons.•Dynamic responses depend on neuron electrophysiological properties and support various computations.•Various modes of dendritic input summation enrich neuron computational repertoires.•Synaptic dynamics further provides differential and frequency-dependent signalling.

At the single neuron level, information processing involves the transformation of input spike trains into an appropriate output spike train. Building upon the classical view of a neuron as a threshold device, models have been developed in recent years that take into account the diverse electrophysiological make-up of neurons and accurately describe their input-output relations. Here, we review these recent advances and survey the computational roles that they have uncovered for various electrophysiological properties, for dendritic arbor anatomy as well as for short-term synaptic plasticity.