期刊名称：International Journal of Computer Science and Network Security
印刷版ISSN：1738-7906
出版年度：2006
卷号：6
期号：12
页码：73-83
出版社：International Journal of Computer Science and Network Security
摘要：We developed a methodological approach to genetic class discovery using gene expression microarray data, which is based a on statistically-oriented class-prediction method called Statistically Weighted Voting (SWV) analysis integrating with clinical risk factor and survival analyses, and statistics of Gene Ontology annotation terms which we use to validate candidate biomarker selection. Our approach provides a ""voting"" class prediction function constructed using the most informative and robust discrete segments (sub-regions) of all covariate ranges and their gradated pairs, which thus allows to model the interactions of variables (genes). We show here that the SWV-based methodology can be adapted for microarray data and profitably used to biomarker selection and discovered two genetic classes associated with essentially improvement of classical histological grade II of human breast cancer. Our findings show that small and reliable genetic grade signatures could improve an individual prognosis for patients with histologic grated II and, thus after further biomedical validation, be used in therapeutic planning for breast cancer patients.
关键词：Voting Algorithms, Biomarker Selection, Prediction, Microarray, Histologic Grades, Cancer Classification.
