| cDNA芯片缺失值处理对基于基因表达谱的疾病分类的影响 |
| |
| 引用本文： | 王栋,郭政,李霞,吕莹丽,朱晶,王晨光.cDNA芯片缺失值处理对基于基因表达谱的疾病分类的影响[J].高技术通讯,2006,16(5):501-505. |
| |
| 作者姓名： | 王栋 郭政 李霞 吕莹丽 朱晶 王晨光 |
| |
| 作者单位： | 1. 哈尔滨医科大学生物信息学系,哈尔滨,150086
2. 哈尔滨医科大学生物信息学系,哈尔滨,150086;同济大学生命科学与技术学院,上海,200092;哈尔滨基太生物芯片开发有限责任公司,哈尔滨,150090 |
| |
| 摘 要： | 选取了4套cDNA芯片数据,分别运用补零和K近邻的方法,对有检测缺失的基因进行了补缺失值处理,分析了不同处理对支持向量机、K近邻分类器、决策树三种分类器分类效能的影响.结果显示: 在cDNA基因表达谱数据中,对检测缺失率不高于5%的基因补缺失值是一种较好的策略,这样可以保留较多的基因供后续的功能分析,同时仍然能够保持很高的疾病分类效能.
|
| 关 键 词： | 基因表达谱 缺失值 分类 |
| 收稿时间： | 2005-06-06 |
| 修稿时间： | 2005年6月6日 |
Influence of missing values replacement on disease classification analysis based on gene expression profiles |
| |
| Wang Dong,Guo Zheng,Li Xia,Lv Yingli,Zhu Jing,Wang Chenguang.Influence of missing values replacement on disease classification analysis based on gene expression profiles[J].High Technology Letters,2006,16(5):501-505. |
| |
| Authors: | Wang Dong Guo Zheng Li Xia Lv Yingli Zhu Jing Wang Chenguang |
| |
| Affiliation: | Department of Bioinformatics, Harbin Medical University, Harbin 150086; School of Biological Science and Technology, Tongji University, Shanghai 200092; Harbin Gene-Tech Biochip Development Inc. ,LTD, Harbin 150090 |
| |
| Abstract: | In this article, two different missing value treatments (replacing with zeros, KNN estimations) combined with three kinds of classifiers,support vector machine(SVM), K-nearest neighbor(KNN) and decision tree(DT), were used to evaluate the effect on four data sets.The results showed that when the missing value rate was less than 5%, enough genes for classification will remain and quite high classification accuracy can be still got. |
| |
| Keywords: | gene expression profile missing value classification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
