Constructing gene expression search engine with partial rank information

近年来在基因综合表达观察方面的研究进展，如微阵列技术，产生了许多实验结果。目前微阵列数据库记录的实验数量已超过40万。数据库通过关键词提供搜索界面，但由于实验科学家和用户的关键词不同，很难选择合适的关键词。为了克服这些问题，我们研究了(1)一种从用户给定的基因表达中搜索基因表达及其功能的方法(2)一个搜索最近基因表达的网站的实现。

Recent progress for observation of comprehensive gene expression such as microarray produces many experimental results. The number of records in microarray database is now more than 400,000 experiments. The database provides the search interface by using keywords, but it is difficult to select proper keywords because of the difference of keywords between experimental scientists and users. To overcome these problems, we studied (1) a search method to characterize gene expression with their functions from user-given gene expressions (2) an implementation of the Web site to search the nearest gene expressions.