NeuroElectro: A database describing the electrophysiology properties of different neuron types

Brains achieve efficient function through implementing a division of labor, in which different neurons serve distinct computational roles. One of the most striking ways in which neuron types differ is in their electrophysiology properties. These properties arise through combinations of ion channels that collectively define the computations that a neuron performs on its inputs. Though the electrophysiology of many neuron types has been previously characterized, these data exist across thousands of journal articles, making cross-study neuron-to-neuron comparisons difficult. Here, using a combination of manual and automated methods, we describe a methodology to curate neuron electrophysiology information into a centralized database.

We developed methods to extract neuron electrophysiology information from formatted data tables contained within the journals Journal of Neuroscience and Journal of Neurophysiology, which contain the majority of the this published information. Using web searching and html parsing tools in Python, we found and stored 1600 electrophysiology data tables across 500 articles. Because authors often use different terms to refer to the same electrophysiology concept, e.g. "Vrest" and "RMP" both refer to a neuron's resting membrane potential, we found the need to develop a basic electrophysiology ontology. Similarly, we used Neurolex's existing neuron ontology (http://neurolex.org) to map different terminology to neuron concepts. We validated our automated methods through manual inspection of a subset of the data. While electrophysiology concept identification was highly reliable (>80%), we found that identifying the correct neuron was less accurate (<50%), in part because of the incompleteness of the neuron ontology, suggesting the need for a two-stage automatic and manual approach.

We hope that this database will be of use to those interested in validating their own measurements on neuron electrophysiology. Furthermore, we plan to integrate this database with existing neuron databases on morphology or gene expression. We hope that these databases will lead to a quantitative understanding of the computational function of different neuron types.