Data Mining of Gene Arrays for Biomarkers of Survival in Ovarian Cancer.

ABSTRACT: The expected five-year survival rate from a stage III ovarian cancer diagnosis is a mere 22%; this applies to the 7000 new cases diagnosed yearly in the UK. Stratification of patients with this heterogeneous disease, based on active molecular pathways, would aid a targeted treatment improving the prognosis for many cases. While hundreds of genes have been associated with ovarian cancer, few have yet been verified by peer research for clinical significance. Here, a meta-analysis approach was applied to two carefully selected gene expression microarray datasets. Artificial neural networks, Cox univariate survival analyses and T-tests identified genes whose expression was consistently and significantly associated with patient survival. The rigor of this experimental design increases confidenc

SUBMITTER: Coveney C 

PROVIDER: S-EPMC4996375 | biostudies-literature | 2015 Jul

REPOSITORIES: biostudies-literature