估計基因網路及基因間函數結構之統計方法

Abstract

我們提出一個利用統計方法來建構基因網路的方法。我們將目標分成三個部分。第一部分是先探討兩個變數間是否有關係存在？如果他們之間確實有關係，第二部分就是探討他們是如何的相關？激發或是抑制?接著第三步即是去找出他們相關的方向。在資料分析的部分，利用了蒙地卡羅模擬分析來展示我們所提出方法的分析效率，並且最後我們也把提出的方法應用在酵母菌資料上來作個討論。

We propose a procedure for constructing gene networks using statistical methods. Three issues are considered. First, is there a relationship between a pair of genes? Second, how do they relate, repress or activate? Third, what is the related direction? By considering the relationship of a pair of genes at a time, our method gives not only the relation (activate, repress) but also the direction for each pair of genes. We conduct Monte Carlo simulations to show the effectiveness of the proposed method. Finally, the method is applied to the Sacharomyces cerevisiae gene expression data as an illustrative example.