NineML: the network interchange for ne uroscience modeling language

Objective Like all scientific research, computational neuroscience research must be reproducible. Big data science, including simulation research, cannot depend exclusively on journal articles as the method to provide the sharing and transparency required for reproducibility. Methods Ensuring model reproducibility requires the use of multiple standard software practices and tools, including version control, strong commenting and documentation, and code modularity. Results Building on these standard practices, model sharing sites and tools have been developed that fit into several categories: 1. standardized neural simulators, 2. shared computational resources, 3. declarative model descriptors, ontologies and standardized annotations; 4. model sharing repositories and sharing standards. Conclusion A number of complementary innovations have been proposed to enhance sharing, transparency and reproducibility. The individual user can be encouraged to make use of version control, commenting, documentation and modularity in development of models. The community can help by requiring model sharing as a condition of publication and funding. Significance Model management will become increasingly important as multiscale models become larger, more detailed and correspondingly more difficult to manage by any single investigator or single laboratory. Additional big data management complexity will come as the models become more useful in interpreting experiments, thus increasing the need to ensure clear alignment between modeling data, both parameters and results, and experiment.

show abstract

“…In NineML the levels of description are not physical scales but are more conceptual, with an abstraction layer and a user layer [108]. …”

Section: Declarative Model Descriptionsmentioning

confidence: 99%

Reproducibility in Computational Neuroscience Models and Simulations

McDougal

Bulanova

Lytton

2016

IEEE Trans. Biomed. Eng.

View full text Add to dashboard Cite

show abstract

“…NeuroML is a model definition languages that is suitable to describe neuronal network models, providing a description of the underlying physiology (Gleeson et al, 2010). NineML is targeted at the simulator independent description of neuron and synapse models as well as connection routines that may then provide the primitives to define network models in NeuroML (Raikov et al, 2011). These efforts are crucial to facilitate collaborations and sharing of models within the community.…”

Section: Discussionmentioning

confidence: 99%

Supercomputers Ready for Use as Discovery Machines for Neuroscience

Helias¹,

Kunkel²,

Masumoto³

et al. 2012

Front. Neuroinform.

Self Cite

View full text Add to dashboard Cite

NEST is a widely used tool to simulate biological spiking neural networks. Here we explain the improvements, guided by a mathematical model of memory consumption, that enable us to exploit for the first time the computational power of the K supercomputer for neuroscience. Multi-threaded components for wiring and simulation combine 8 cores per MPI process to achieve excellent scaling. K is capable of simulating networks corresponding to a brain area with 108 neurons and 1012 synapses in the worst case scenario of random connectivity; for larger networks of the brain its hierarchical organization can be exploited to constrain the number of communicating computer nodes. We discuss the limits of the software technology, comparing maximum filling scaling plots for K and the JUGENE BG/P system. The usability of these machines for network simulations has become comparable to running simulations on a single PC. Turn-around times in the range of minutes even for the largest systems enable a quasi interactive working style and render simulations on this scale a practical tool for computational neuroscience.

show abstract

“…Different approaches are used to simplify the interfaces to existing simulators which offer different trade-offs in the ease of reusability of components, the complexity of the surrounding toolchains, and in the expressiveness and readability of the model specifications. One approach has been to design high-level, declarative languages for expressing the mathematical behaviors of neurons and synapses, for example MODL (Carnevale and Hines, 2006), NDF (Bower and Beeman, 1998), NeuroML/LEMS (Goddard et al, 2001; Gleeson et al, 2010) and NineML (Raikov et al, 2011). These languages are either read natively by the simulator, or intermediate code-generation tools can be used to translate these descriptions to run on existing simulators.…”

Section: Introductionmentioning

confidence: 99%

morphforge: a toolbox for simulating small networks of biologically detailed neurons in Python

Hull

Willshaw²

2014

Front. Neuroinform.

View full text Add to dashboard Cite

The broad structure of a modeling study can often be explained over a cup of coffee, but converting this high-level conceptual idea into graphs of the final simulation results may require many weeks of sitting at a computer. Although models themselves can be complex, often many mental resources are wasted working around complexities of the software ecosystem such as fighting to manage files, interfacing between tools and data formats, finding mistakes in code or working out the units of variables. morphforge is a high-level, Python toolbox for building and managing simulations of small populations of multicompartmental biophysical model neurons. An entire in silico experiment, including the definition of neuronal morphologies, channel descriptions, stimuli, visualization and analysis of results can be written within a single short Python script using high-level objects. Multiple independent simulations can be created and run from a single script, allowing parameter spaces to be investigated. Consideration has been given to the reuse of both algorithmic and parameterizable components to allow both specific and stochastic parameter variations. Some other features of the toolbox include: the automatic generation of human-readable documentation (e.g., PDF files) about a simulation; the transparent handling of different biophysical units; a novel mechanism for plotting simulation results based on a system of tags; and an architecture that supports both the use of established formats for defining channels and synapses (e.g., MODL files), and the possibility to support other libraries and standards easily. We hope that this toolbox will allow scientists to quickly build simulations of multicompartmental model neurons for research and serve as a platform for further tool development.

show abstract

NineML: the network interchange for ne uroscience modeling language

Cited by 29 publications

References 0 publications

Reproducibility in Computational Neuroscience Models and Simulations

Reproducibility in Computational Neuroscience Models and Simulations

Supercomputers Ready for Use as Discovery Machines for Neuroscience

morphforge: a toolbox for simulating small networks of biologically detailed neurons in Python

Contact Info

Product

Resources

About