作者: M. G. Kounelakis , M. E. Zervakis , G. C. Giakos , X. Kotsiakis
DOI: 10.1007/978-3-642-23508-5_98
关键词:
摘要: The aim of this study is to reveal the significance glycolysis-related genes in design new brain gliomas treatment protocols. Towards direction a feature selection and classification method, embedding Relief-F filter criterion under Support Vector Machines (SVM) classifier, has been introduced focusing on identification significant genetic alterations related glycolysis. In particular genomic (Microarray Expression) dataset, consisting 14 glioma patients used for statistical analysis. results have demonstrated that specific group gene markers (HK, PGI, PFK, ALDO, GAPDH, PGK, PGM, ENO, PK, LDH, PDH MDH) directly cell glycolysis, great impact discrimination different grades malignancy (AUROC 0.98), thus leading conclusion these could establish foundations therapeutic approaches.