The aim of this project is to expand the scope of PyNN from networks of simplified neuron models, such as integrate-and-fire, to networks consisting of biophysically- and morphologically-detailed neuron models, targeting the NEURON and Arbor simulators, the SpiNNaker and BrainScaleS neuromorphic computing systems, and interoperability with NeuroML and the SONATA format.

Job Details:

General Information

•  Reference : UMR9197-ODILEC2-078
• Reference : UMR9197-ANDDAV-007
• Workplace : GIF SUR YVETTE
• Date of publication : Thursday, June 24, 2021
• Type of Contract : FTC Scientist
• Contract Period : 18 months
• Expected date of employment : 1 September 2021
• Proportion of work : Full time
• Remuneration : between 2728 and 4323 € gross per month, depending on experience
• Desired level of education : PhD
• Experience required : Indifferent

Missions:

A position in neuronal network modelling and simulation is available within the Neuroinformatics research group of the Paris-Saclay Institute of Neuroscience, as part of the Human Brain Project.

The large number of different simulation environments in neuroscience presents both a problem and an opportunity. Since each simulator has its own programming interface and output data format, reproducing, reusing, or combining models written for different simulators typically presents considerable difficulty. On the other hand, since each simulation environment has a different balance between efficiency, flexibility, scalability and user-friendliness, we can choose the most appropriate one for a given problem, or cross-check the results obtained with different tools.
Simulator-agnostic model representation languages such as NeuroML and PyNN allow us to preserve the advantages of simulator diversity while reducing or eliminating the disadvantages, thereby making reproducibility and reuse much easier. Such languages let us take a single model description and simulate it in different environments without having to reimplement the model each time.
The aim of this project is to expand the scope of PyNN from networks of simplified neuron models, such as integrate-and-fire, to networks consisting of biophysically- and morphologically-detailed neuron models, targeting the NEURON and Arbor simulators, the SpiNNaker and BrainScaleS neuromorphic computing systems, and interoperability with NeuroML and the SONATA format.

Activities:

• Extend the PyNN API to multicompartmental neuron models, incorporating biophysical models of ion channels, synaptic mechanisms and intracellular calcium dynamics, building on existing prototypes and in consultation with the wider modelling community
• complete the implementation of this API for the NEURON and Arbor simulators
• develop input/output support for the NeuroML and SONATA formats
• develop appropriate unit and system tests
• write documentation and tutorials, and participate in user training activities
• publish on topics of modelling/simulation technologies, workflows and model sharing/reuse

Skills:

• PhD in neuroscience, physics, computer science, or a closely related discipline
• preferably with experience of computational modelling.
•  experience of simulation tools often used in computational neuroscience, such as NEURON, NEST or Brian
•  knowledge of computer programming (in Python, C++, MATLAB or other)
•  knowledge of French would be helpful but is not required, as English is the working language of the project.

Work Context:

The Paris-Saclay Institute of Neuroscience is a research unit of the CNRS and the University Paris-Saclay, located in Gif sur Yvette, about 40 minutes from central Paris. The Neuroinformatics research group develops novel informatics tools and approaches for neuroscience data sharing, modelling and simulation, and for brain-inspired computing. This post is funded by the Human Brain Project, an EU Flagship initiative in which over 100 partners work together to build a completely new information computing technology infrastructure for neuroscience and for brain-related research in medicine and computing.

Contact:

Andrew Davison

Please click here to learn more.