New release of NeuroML 2 and LEMS and two recent publications

[Padraig Gleeson]

Dear colleagues,

We are happy to announce a new stable release (version 2, beta 3) of the NeuroML language for model specification in computational neuroscience.

http://www.neuroml.org/getneuroml

NeuroML allows specification of models of neuronal systems from integrate and fire cells up to complex 3D networks of multicompartmental neurons. NeuroML version 2 has been extensively redesigned to be built on a new language, LEMS, which allows machine readable definitions of model structure and dynamics. This facilitates model transparency, portability and code generation. LEMS and NeuroML 2 are described in detail in:

Robert C. Cannon, Padraig Gleeson, Sharon Crook, Gautham Ganapathy, Boris Marin, Eugenio Piasini and R. Angus Silver, LEMS: A language for expressing complex biological models in concise and hierarchical form and its use in underpinning NeuroML 2, Frontiers in Neuroinformatics 2014, doi: 10.3389/fninf.2014.00079

There are libraries for reading, writing and simulating NeuroML models in Java (https://github.com/NeuroML/jNeuroML) and Python. The Python APIs for NeuroML and LEMS have recently been described in:

Michael Vella, Robert C. Cannon, Sharon Crook, Andrew P. Davison, Gautham Ganapathy, Hugh P. C. Robinson, R. Angus Silver and Padraig Gleeson, libNeuroML and PyLEMS: using Python to combine procedural and declarative modeling approaches in computational neuroscience Frontiers in Neuroinformatics 2014, doi: 10.3389/fninf.2014.00038

The growing number of tools, libraries and databases supporting both NeuroML v1 and NeuroML v2 are listed here:

http://www.neuroml.org/tool_support

Open Source Brain is a repository of models in computational neuroscience which actively supports open, collaborative development of models, as well as conversion to simulator independent formats including NeuroML. An overview of the models currently present and the formats supported can be found here:

http://www.opensourcebrain.org/status

Regards,
The NeuroML development community

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.gl...@ucl.ac.uk
----------------------------------------------------- 

[Padraig Gleeson]

Hi,

A quick update on what's been happening in NeuroML 2 development since the publication of the LEMS/NeuroML2 paper & release of NeuroML v2 beta 3 (see mail below).

- The main NeuroML website has been updated. There is also now a single page linking to all the latest specifications, examples, documentation, core tools, etc.: http://www.neuroml.org/getneuroml.

- A question we're frequently asked is "How do I convert my model to NeuroML?". There is now a page on OSB dedicated to this: http://www.opensourcebrain.org/docs#Converting_To_NeuroML2

- Developments towards NeuroML2 beta 4 release have centred on implementation of support for gap junctions (and analogue synapses) in NeuroML2. Support for these have required updates to LEMS (so they can be simulated natively using jLEMS), as well as the mappings to simulator formats, particularly NEURON. Examples of NeuroML2 networks with gap junctions can be found in some of the OpenWorm example simulations (e.g. here and here).

- jNeuroML has been constantly updated with improved support for export to a number of different formats (try: jnml -?), including generating scripts for NEURON for (multicompartmental) cell and network models specified in NeuroML2, and generation of images of cells & networks in SVG format (jnml MyNet.nml -svg)

- pyNeuroML - is intended as a single Python package for handling NeuroML 2 which wraps the functionality present in jNeuroML to make it accessible in Python scripts. Still a work in progress, but the examples should highlight some of the current features. It also is a home for a number of other Python based tools for handling NeuroML 2 including exporting cell morphologies from NEURON and analysis of channels in NeuroML & comparison to the corresponding NMODL files. Examples of both these features are referenced here.

- A core set of models on OSB have been identified as candidates for "best practices" examples of NeuroML 2 (and PyNN). These will be the projects we will focus on over the next few months to make sure they are as compliant as possible to NeuroML 2 and highlight the best use of associated tools (e.g. OSB 3D Explorer) and libraries. See https://github.com/OpenSourceBrain/redmine/issues/125. There is also a list of curated/validated examples in NeuroML v1.8.1: https://github.com/OpenSourceBrain/redmine/issues/119


There will be another NeuroML Editorial Board meeting on 11th May before the Open Source Brain meeting in Sardinia. Here is an outline of some of the topics up for discussion:

1) Review of NeuroML/LEMS progress since last year
2) NeuroML 2 Release beta 4
3) LEMS development - Java and Python support   
4) NeuroML website/NeuroML DB  
5) Editorial board process - elections 2015
6) Relationship with NineML and other initiatives
7) Future developments of NeuroML

If there are any other topics which any of the community would like us to discuss, please get in contact.

Regards,
Padraig