LEMS: a language for expressing complex biological models in concise and hierarchical form and its use in underpinning NeuroML 2

Abstract

Computational models are increasingly important for studying complex neurophysiological systems. As scientific tools, it is essential that such models can be reproduced and critically evaluated by a range of scientists. However, published models are currently implemented using a diverse set of modeling approaches, simulation tools, and computer languages making them inaccessible and difficult to reproduce. Models also typically contain concepts that are tightly linked to domain-specific simulators, or depend on knowledge that is described exclusively in text-based documentation. To address these issues we have developed a compact, hierarchical, XML-based language called LEMS (Low Entropy Model Specification), that can define the structure and dynamics of a wide range of biological models in a fully machine readable format. We describe how LEMS underpins the latest version of NeuroML and show that this framework can define models of ion channels, synapses, neurons and networks. Unit handling, often a source of error when reusing models, is built into the core of the language by specifying physical quantities in models in terms of the base dimensions. We show how LEMS, together with the open source Java and Python based libraries we have developed, facilitates the generation of scripts for multiple neuronal simulators and provides a route for simulator free code generation. We establish that LEMS can be used to define models from systems biology and map them to neuroscience-domain specific simulators, enabling models to be shared between these traditionally separate disciplines. LEMS and NeuroML 2 provide a new, comprehensive framework for defining computational models of neuronal and other biological systems in a machine readable format, making them more reproducible and increasing the transparency and accessibility of their underlying structure and properties.

Open Source Brain: a collaborative resource for visualizing, analyzing, simulating and developing standardized models of neurons and circuits

“…1a ). These features, which expose the properties of models in a transparent manner, were made possible by defining models in the structured model specification languages NeuroML 26,27 or PyNN 31 , since these can be parsed by specialized software to extract and visualize key aspects of the model (e.g. biophysical parameters, cell morphology) and transform them into a format that is familiar to neuroscientists.…”

“…In contrast, modular programs with standardized building blocks can be more easily tested with automated routines and then assembled into larger structures, minimizing errors. In neuroscience, standardized model descriptions, such as NeuroML 26,27 , provide a defined structure that could be used for such a modular approach. Moreover, large-scale model development is beginning to involve groups of developers 28 using version control software such as Git 29 , practices that are common in industry, but remain the exception rather than the rule in academia 30 .…”

“…NeuroML is a widely used standard for declarative model description in neuroscience. NeuroML version 2 (the version we refer to in this paper) is built on a flexible low-level language, LEMS (low entropy model specification), that enables a wide range of physico-chemical processes to be defined 26 . The NeuroML2/LEMS framework allows diverse membrane conductances (Hodgkin-Huxley 5 , kinetic scheme, or any other formalism based on ordinary differential equations and state-dependent events), synaptic models (including chemical with short and long-term plasticity, electrical or analog/graded), point neurons or detailed neuronal morphologies and circuit connectivity to be specified and mapped to supported simulators ( Table 1 ).…”

Temporal constrained objects for modelling neuronal dynamics

“…NineML (Raikov, 2010) uses XML based abstract object models and enable quick prototyping of neuron, synapse and network models using parameters for model variables, state update rules and mathematical descriptions. Low entropy model specification also follows a similar approach and are more flexible in defining and translating models (Cannon et al, 2014). Even though the XML-based model description frameworks reduce external software dependencies, they do not provide any details on how to simulate the models.…”

Specifications of Standards in Systems and Synthetic Biology: Status and Developments in 2016

“…NeuroML, a model description language for computational neuroscience [24]. The current specification is NeuroML version 2, beta 4, and can be found on the NeuroML website [25].…”