Efficient Simulation of Cell-Biological Multi-Level Models (ESCeMMo)

在细胞生物学中，支持多层次建模和模拟的更具表现力的语言的数量正在稳步增长。它们的丰富功能增加了复杂动态系统建模的灵活性和简便性，但也给高效的仿真带来了挑战。ESCeMMO(细胞生物多层模型的高效模拟)研究项目的目的是开发更高效的执行策略，以考虑可表达的多层语言的特性和要求。由于并非所有模型都将平等地使用表达语言提供的全套功能，因此应自动选择和应用利用特定约束进行更有效模拟的量身定制的模拟器。应开发多种算法方法，其中包括对子模型的有效划分，以及适当模拟器的选择和同步。应探索模型分解和功能分解策略，以便更有效地并行化单个和多个仿真运行。这项研究将基于ML-Rules，这是一种基于规则的表达语言，用于细胞生物学中的多层次建模和模拟。开发的方法将通过性能研究和对线粒体的模拟研究进行评估。后者还将利用建模语言的主要功能。该项目的结果将增加模拟多层次模型的效率，丰富可表达建模语言的执行方法组合，并将增加它们在回答细胞生物学研究问题方面的适用性。

The number of more expressive languages that support a multi-level modeling and simulation is steadily growing in cell biology. Their rich features add to the flexibility and ease in modeling complex dynamic systems, but they also pose a challenge for efficient simulation. The aim of the research project ESCeMMo (Efficient Simulation of Cell-biological Multi-Level Models) is to develop more efficient execution strategies that take properties and requirements of expressive multi-level languages into account. As not all models will equally use the full set of features offered by expressive languages, tailored simulators that exploit specific constraints for more efficient simulation shall automatically be selected and applied. Multi-algorithm approaches shall be developed, which comprise an effective partitioning into sub-models, and the selection and synchronization of suitable simulators. Model decomposition as well as functional decomposition strategies shall be explored, for more effectively parallelizing single and multiple simulation runs. This research will be based on ML-Rules, an expressive rule-based language for multi-level modeling and simulation in cell biology. The developed methods will be evaluated by performance studies and a simulation study on mitochondria. The latter will also exploit central features of the modeling language. The results of the project will add to the efficiency of simulating multi-level models, enrich the portfolio of execution methods for expressive modeling languages, and will increase their applicability for answering cell biological research questions.