Oxford University Computing Laboratory, University of Oxford, Wolfson Building, Parks Road, Oxford OX13QD, UK.
Prog Biophys Mol Biol. 2011 Oct;107(1):11-20. doi: 10.1016/j.pbiomolbio.2011.06.003. Epub 2011 Jun 15.
Effective reuse of a quantitative mathematical model requires not just access to curated versions of the model equations, but also an understanding of the functional capabilities of the model, and the advisable scope of its application. To enable this "functional curation" we have developed a simulation environment that provides high-throughput evaluation of a mathematical model's functional response to an arbitrary user-defined protocol, and optionally compares the results against experimental data. In this study we demonstrate the efficacy of this simulation environment on 31 cardiac electrophysiology cell models using two test cases. The S1-S2 response is evaluated to characterise the models' restitution curves, and their L-type calcium channel current-voltage curves are evaluated. The significant variation in the response of these models, even when the models represent the same species and temperature, demonstrates the importance of knowing the functional characteristics of a model prior to its reuse. We also discuss the wider implications for this approach, in improving the selection of models for reuse, enabling the identification of models that exhibit particular experimentally observed phenomena, and making the incremental development of models more robust.
有效复用定量数学模型不仅需要访问经过策展的模型方程版本，还需要了解模型的功能能力和其应用的建议范围。为了实现这种“功能策展”，我们开发了一个模拟环境，该环境可以对数学模型对任意用户定义协议的功能响应进行高通量评估，并可选择将结果与实验数据进行比较。在这项研究中，我们使用两个测试用例在 31 个心脏电生理学细胞模型上演示了这种模拟环境的功效。通过评估 S1-S2 响应来描述模型的恢复曲线，并评估它们的 L 型钙通道电流-电压曲线。即使模型代表相同的物种和温度，这些模型的响应也存在显著差异，这表明在重复使用模型之前了解模型的功能特征非常重要。我们还讨论了这种方法的更广泛意义，即可以改进模型的复用选择，能够识别出表现出特定实验观察现象的模型，并使模型的增量开发更加稳健。
