Foundations of Computational Knowledge Discovery from cDNA Microarray Data

自本研究项目开始以来，cDNA微阵列已被广泛用于实验室中基因表达的测量。一些全基因组基因表达分析项目已经计划将商业化的统计分析程序包和数据库作为生物信息学工具，以特别的方式使用。然而，在日本，从实践的角度来看，基于信息科学的基础并没有得到足够的重视，只是开发了软件工具。该研究项目促进了新型微阵列数据分析技术的发展。首先，我们从cDNA微阵列数据创建了一个基因网络模型的数学框架（布尔网络模型、贝叶斯网络模型、常微分方程系统），然后我们开发了从cDNA微阵列数据推断这些模型的计算方法。这些方法都是通过真实的生物分析来检验的。第二个贡献是基因网络建模和仿真技术的发展（混合Petri网和各种生物途径模型仿真的延伸）。有了这些研究成果，我们已经为发展系统生物学建立了一个计算策略。

Since the beginning of this research project, the cDNA microarrays have been intensively employed for measuring gene expressions in laboratories. Some projects on genome-wide gene expression analysis have planed where commercialized statistical analysis program packages and databases are employed as bioinformatics tools in an ad hoc way. However, foundations based on information science were not paid attentions so much while only software tools have been developed from the viewpoint of practice in Japan.This reseach project has contributed to the development of novel microarray data analysis technology. First, we have created a mathematical framework for gene network models from cDNA microarray data (Boolean network model, Bayesian network model, system of ordinary differential equations), then we have developed computational methods for inferring these models from cDNA microarray data. These methods were examined through real biological analyzes. The second contribution is a development of modeling and simulation technology for gene networks (an extention of hybrid Petri net and model simulation of various biopathways).With these research contributions, we have established a computational strategy for developing systems biology.