基于数据处理的肿瘤基因选择系统  

Research on tumor gene selection system based on data processing

作　　者：田梓君 崔新于 

机构地区：[1]河南师范大学计算机与信息工程学院,河南新乡453007

出　　处：《无线互联科技》2017年第8期35-36,共2页Wireless Internet Technology

基　　金：河南师范大学大学生创新创业训练计划校级立项项目;项目编号:20150016

摘　　要：近年来,随着肿瘤医院就诊人数的不断增多,特征基因提取已成为中内外学者研究的热门,研究成果也为临床癌症的分析诊断及预测提供了极大的便利。然而,由于基因表达谱数据具有维度高、样本少、复杂多样的特点,准确地挖掘基因数据中所蕴含的肿瘤信息基因成为当前的首要挑战。文章阐述了在Matlab2010b编程环境下开发的肿瘤基因数据选择系统,介绍了改进后信噪比与支持向量机回归特征消去(SVM-RFE)的基因选择方法相结合对于基因选择的优异性,对于筛选出对分类有益的特征基因具有良好的效果。In recent years, with the increasing number of people who go to tumor hospitals, the extraction of characteristic genes hasbecome a hot topic in domestic and foreign scholars. The research results have also provided great convenience for the diagnosis andprediction of clinical cancers. However, the gene expression data has the characteristics of high dimension, less sample, complex anddiverse, it is the most important challenge to accurately dig the tumor information gene contained in gene data. In this paper, the selectionof tumor gene data selection system was developed in Matlab2010b programming environment, and the advantages of improved signal-to-noise ratio and support vector machine regression feature elimination （SVM-RFE） gene selection method are also introduced. This systemhas a good effect on screening out the characteristic genes that are beneficial to the classification.

关 键 词：特征基因 系统开发 改进的信噪比 

分 类 号：R730.4[医药卫生—肿瘤] TP18[医药卫生—临床医学]