Resumo: |
A decade ago, the creation of the Systems Biology Markup Language (SBML) changed the way people exchanged, verified and re-used models in systems biology. The robustness and versatility of this format, coupled to a wide software support, fostered the emergence of an entire area of research centered on model processing such as encoding, annotation, merging, comparison and integration with other datasets. Recently, new languages appeared that complement the model description, such as SED-ML to describe the simulation experiments or SBRML to encode the numerical results. In neuroscience, other fledgling efforts cover for instance multi-compartment neurons with NeuroML, and neuronal networks with NineML. More are needed to cover the wide spectrum of computational models used in neuroscience. The developers of those initiatives are in contact, and try to improve the interoperability of the languages, for instance by sharing metadata. Similar development guidelines, governance principles and quality checks are needed, in order to provide the community with a serious infrastructure. One can hope to see, in a not too elusive future, the creation of a coherent set of non-overlapping standards that will support not only the various modeling approaches and scales needed to simulate human functions and dysfunctions, but also cover model structure, parametrization, simulation and numerical output. Such a toolkit will allow the bridging of genomics, computational neuroscience and drug discovery.
|