Derivation of a protocol agnostic current clamp description for the cross-analysis of diverse neuronal electrophysiological databases

Biophysically realistic spiking network models rely on the ability of the cellular models to be coherent with the biological diversity of cellular properties and electrophysiological behavior, for example how synaptic conductance inputs impact neuronal firing, Numerous studies of neuronal biophysics have been made over the last decades, primarily intracellular recordings under the current clamp configuration, thus current input as an approximation of synaptic inputs. Increasingly labs publicly release their experimental databases, often including other cellular information [1,2,3]. Apart from common standards for electrophysiological data (Neurodata Without Borders, [4]), experimental design and data collation is typically unique to a given lab, complicating comparative analysis across databases. To overcome this limitation, we propose a database-independent analysis pipeline to analyze intracellular recording data and extract features relevant for neuron models. These features constitute a Protocol-Agnostic Current-Clamp (PACC) vector, characteristic of a cell independently of its database. Examples of the application of the PACC include associating transcriptomic or anatomic data available in one database, to cells without this information from another database. Another application is to associate current-clamp and conductance-clamp data from our lab, allowing the inference of current-clamp data from other labs to conductance dependent properties.