Meta – Analysis of Genomic Data in Endometrial Cancer

Abstract

Endometrial Carcinoma (EC) is the most common gynecologic malignancy in the Asian Countries as well as other developing countries. The majority of cases are diagnosed when the carcinoma is confined to the uterus, leading to less than 1.5% of cancer deaths. Endometrial Cancers largely caused by hormonal imbalances. But the research done to identify the genomic level impact is limited. Among different types of mutations Copy Number Variation(CNV) has a significant effect in causing tumors. Studying the Copy Number Variation (CNV) in hormone responsive genes of endometrial cancer would help to understand the biological underpinning of endometrial cancer progression and to determine the treatment strategies. In this study, we have proposed an approach to identify the genomic therapies for endometrial cancer by predicting the gene alteration associated with hormone responsive genes. Patient's clinical data are analyzed and identified how the alterations of hormone responsive genes. In this study Classification is used to predict the correlation between patient data samples and classes by analyzing Patients data. Prediction model is built using Support Vector Machine (SVM) to discover correlation between patient clinical data and gene alterations. In this study we used two approaches to build the prediction model. Predicting the variation of single gene and predicting the variation of gene profiles are two methodologies. Output of the study shows that, the proposed SVM model gives a strong and an accurate prediction for possible alteration of hormone responsive gene of a patient.