SBML to bond graphs: from conversion to composition

Abstract

The Systems Biology Markup Language (SBML) is a popular software-independent XML-based format for describing models of biological phenomena. The BioModels Database is the largest online repository of SBML models. Several tools and platforms are available to support the reuse and composition of SBML models. However, these tools do not explicitly assess whether models are physically plausibile or thermodynamically consistent. This often leads to ill-posed models that are physically impossible, impeding the development of realistic complex models in biology. Here, we present a framework that can automatically convert SBML models into bond graphs, which imposes energy conservation laws on these models. The new bond graph models are easily mergeable, resulting in physically plausible coupled models. We illustrate this by automatically converting and coupling a model of pyruvate distribution to a model of the pentose phosphate pathway.Graphical AbstractHighlightsA framework to convert suitable SBML models of biochemical networks into bond graphs is developed.The framework is applied here to two interconnecting models of metabolism pathways.We automatically integrate the generated bond graph modules.We qualitatively illustrate the functionality of the composed model.

A review of the diverse applications of bond graphs in biology and physiology

“…convergence criteria and time steps. To further enhance the reusability of these models, BG counterparts of the physiological models are emerging, and tools are being developed to aid the automated generation of BG equivalents to existing models [174]. Thus, the BG model constructed by a researcher can be shared in repositories that other researchers can use to build their own more comprehensive models without having to develop the same module from scratch.…”

“…Shahidi et al [172] used SemGem (a software tool for automating the modular composition and decomposition of biosimulation models) to automatically combine CellML, an open standard language based on the XML (Extensible Markup language), [173] models by converting them to analogous BG models. In the following articles [174], they also converted systems biology markup language [175], and CellML models to BGs to apply the same method of composition to generate complex models from existing modules [176]. In this work, the computational models are first converted into BG models and enriched with semantic annotations.…”