Unifying computational and experimental descriptions of cortical networks across scales

An in-depth understanding of how information is processed in cortical networks will require integration and consolidation of diverse experimental data on these circuits along with computational modelling of their dynamical properties at multiple scales. Many datasets on cortical connectivity at the micro-, meso- and macro-scale have been released in recent years. In parallel, computational models have been created which range from spiking network models with point neurons or multicompartmental cells to whole brain models using neural mass representations for cortical areas. However, integrating data and computational models from different groups has been hampered by the lack of standard representations for data and the diversity of simulators for creating models at different scales.

In this talk I will present my work to date on making computational models more accessible and interoperable, through the development of the model description language NeuroML, and with the development of an open platform for sharing and collaboratively developing such models, Open Source Brain (www.opensourcebrain.org). I will also discuss work to unify descriptions of complex connectivity in cortical networks across experimental datasets and models.