Multiscale Modeling

Understanding the brain requires the integration of huge amounts of heterogeneous and complex data collected at multiple levels of investigation. Multiscale computational modeling techniques allow for integration of neuroscience data on several scales, and help link between observable scales, from e.g. ion channels and single cell types to brain function. This integrative function helps in the formulation of experimentally testable hypotheses and the design of more informative experiments, and thereby increases the usefulness of experimental data.

Some of the most important issues facing the multiscale modeling field concern standards and interoperability. The replicability, reusability and portability of many models is limited, in part due to varying practices in the field with incompatible formats for model files and descriptions. There are many active modeling efforts that span a very wide range of scales, and few of them currently have good standards support. The INCF Program on Multiscale Modeling therefore aims to contribute to the development of new standards and the integration of existing efforts in the field.

Model description and NineML

In recognition of the challenges, several community efforts have been launched, among them NeuroML and PyNN, which enable specification and mapping of model components across independent implementations. The development of NineML - a simulator independent description language for network models - was initiated because an explicit mathematical description of model components, independent of simulator implementation details, was seen to be missing. NineML is a proposal of such a representation, intended to bridge a gap in the representation of network models, and as one step closer to a cross-initiative harmonized standard for representing most type of models, including large-scale networks.

NineML has been developed by an INCF Task Force consisting of representatives from relevant projects including the Blue Brain Project, GENESIS, KInNeSS, MOOSE, NEST, NeuroML, NEURON and PyNN. The official NineML web pages are located at the INCF Software Center.

Connecting large-scale simulations with MUSIC

INCF has also supported the development of MUSIC, a standard interface for run-time exchange of data among parallel applications in a cluster environment. It is designed for interconnecting large-scale neuronal network simulators with each other or with other tools.