Department of Environmental Microbiology, UFZ - Helmholtz Centre for Environmental Research, Permoserstrasse 15, D-04318 Leipzig, Germany.
Bioinformatics. 2010 Jul 15;26(14):1788-9. doi: 10.1093/bioinformatics/btq263. Epub 2010 May 25.
Analysing genome-scale in silico models with stoichiometry-based methods is computationally demanding. The current algorithms to compute chemical organizations in chemical reaction networks are limited to small-scale networks, prohibiting a thorough analysis of large models. Here, we introduce a parallelized version of the constructive algorithm to determine chemical organizations. The algorithm is implemented in the Standard C programming language and parallelized using the message passing interface (MPI) protocol. The resulting code can be executed on computer clusters making use of an arbitrary number of processors. The algorithm is parallelized in an embarrassing parallel manner, providing good scalability.
An implementation of the algorithm including source code can be obtained from http://www.minet.uni-jena.de/csb/prj/ot/tools
使用基于化学计量学的方法对基因组规模的计算模型进行分析在计算上是很耗费资源的。目前用于计算化学反应网络中化学组织的算法仅限于小规模网络，这使得对大型模型的全面分析受到限制。在这里，我们引入了一种基于构建算法的并行版本，用于确定化学组织。该算法是用标准 C 编程语言实现的，并使用消息传递接口 (MPI) 协议进行了并行化。生成的代码可以在计算机集群上执行，利用任意数量的处理器。该算法采用尴尬并行的方式进行并行化，具有良好的可扩展性。
该算法的实现包括源代码，可以从 http://www.minet.uni-jena.de/csb/prj/ot/tools 获得。
