Building cortical network models in NeuroML2 using procedural and declarative programming approaches

There is a long history of neuronal simulator-specific modelling in the fields of theoretical neuroscience and neuroinformatics. As a result, most of the single cell and neural network models have low RAPT (reproducibility, accessibility, portability and transparency). This has slowed the progress of using biophysically and anatomically detailed neural network models in understanding the brain function in health and disease. In a period of three months during the GSoC placement, I would like to contribute to the Open Source Brain (OSB) Initiative which aims to increase the RAPT of single cell and neural network models. The project will extend the initial work by the OSB on converting the thalamocortical network model by Traub et al. (2005) to the simulator-independent model format, NeuroML2. Firstly, I propose to annotate and implement the tests for spiking behaviour of Traub model cells which are converted to the NeuroML2 format. Secondly, I propose to create a generic Python-based package which can generate different configurations of cortical networks in the NeuroML2 format and combine unique network components from different cortical network projects.