基于非负矩阵分解和Normal_Matrix的肿瘤基因分类

Classification of tumor gene based on nonnegative matrix decomposition and Normal_Matrix

文中提出了一种结合非负矩阵分解和Normal_Matrix谱分解技术的肿瘤基因分类方法.其分类过程首先是利用fdr_test记分准则粗略除去噪声基因以实现基因表达谱数据的初步降维,进而运用非负矩阵分解萃取基因间的综合属性,通过综合属性构造样本间的Normal_Matrix并对其进行奇异值分解获取表征样本类别属性的谱分量实现肿瘤类型的分类识别.采用三组具有代表性的肿瘤基因表达谱数据进行实验,通过与其他方法的对比,其结果证明了文中方法的可行性和有效性.

This paper proposed a method for the classification of tumor gene expression data based on nonnegative matrix decomposition and Normal_Matrix spectrum decomposition technology. First, use fdr_test scoring criteria to remove noise genes roughly to reduce the dimensions of gene expression data preliminary Next, extract the comprehensive properties between genes using the nonnegative matrix decomposition, then construct the Normal_Matrix between samples based on the comprehensive properties and do singular value decompositionon of it to gain the spectral component which can describe the class attribute of samples. At last, realize the classification of tumor types. Three representative groups of gene expression data are used for test, and the feasibility and effectiveness of this algorithm has been well proved by contrast tests between other methods.