Publication Type : Journal Article
Source : Applied Artificial Intelligence, Taylor & Francis, Volume 35, Number 3, p.206-226 (2020)
Url : https://www.tandfonline.com/doi/abs/10.1080/08839514.2020.1854988
Campus : Kochi
School : School of Computing
Department : Computer Science
Year : 2020
Abstract : ABSTRACTOvarian cancer stands in the sixth position among the most commonly occurring cancers in the world. Because of the high rate of recurrence, this gynecological malignancy seems to be a vital reason behind cancer-related death among women as tumor recurrence stands as an obstacle in ovarian cancer treatment. It is crucial to find those recurrence causing factors in order to plan suitable therapies with high prognostic results. Hence, in this work, a multistage feature selection methodology is proposed to identify key MiRNAs and clinical features for improving the accuracy of ovarian cancer recurrence prediction. MiRNA expression profiles of ovarian cancer patients and their corresponding clinical data were downloaded from the TCGA cancer repository. From 588 MiRNAs, 6 key MiRNAs were selected using the Inheritable Bi-objective Combinatorial Genetic Algorithm (IBCGA) followed by factor analysis. The biological importance of the resultant MiRNAs in cancer and cellular pathways were studied using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis. Further, recurrence prediction was performed using the obtained MiRNA expression profiles and clinical factors, chosen using correlation analysis. The proposed approach using the selected features yielded a prediction accuracy of 91.86% using the XGBoost classifier while the same without feature selection was 76.59%. Compared to previous similar works, this model provides a better result in terms of accuracy and reveals influential MiRNAs in ovarian cancer.
Cite this Research Publication : S. Sujamol, E. R. Vimina, and Krishnakumar, U., “Improving Recurrence Prediction Accuracy of Ovarian Cancer Using Multi-phase Feature Selection Methodology”, Applied Artificial Intelligence, vol. 35, pp. 206-226, 2020.