EBST: An Evolutionary Multi-Objective Optimization Based Tool for Discovering Potential Biomarkers in Ovarian Cancer

Area of Science:

• Biotechnology
• Bioinformatics
• Oncology

Background:

• Ovarian cancer is the leading cause of gynecologic cancer deaths, largely due to diagnostic challenges.
• High-throughput technologies are driving the search for novel tumor biomarkers for early detection.
• MicroRNAs (miRNAs) are increasingly recognized for their potential as diagnostic biomarkers.

Purpose of the Study:

• To introduce a novel computational tool, EBST (Expression-based Biomarker Selection Tool), for identifying microRNAs with biomarker potential in ovarian cancer.
• To develop and validate a method for selecting potent ovarian cancer microRNA biomarkers using a sophisticated optimization algorithm.

Main Methods:

• The EBST tool employs a Modified Multi Objective Imperialist Competitive Algorithm with objective functions for classifier performance, clustering error, and minimum Redundancy Maximum Relevance (mRMR).
• The study utilized the FDR filter for pre-processing and incorporated four l1-SVM classifier performance metrics and one average mRMR ranking as objective functions.
• Eleven specific microRNAs (e.g., hsa-miR-6784-5p, hsa-miR-1228-5p) were identified as potential biomarkers.

Main Results:

• The proposed model demonstrated exceptional classification performance: 100% sensitivity, 99.38% specificity, 99.69% accuracy, and 99.39% positive predictive value.
• The identified microRNAs were biologically validated using bioinformatics tools and literature review, confirming their involvement in cancer signaling pathways.
• The EBST method outperformed existing state-of-the-art techniques in microRNA biomarker selection for ovarian cancer.

Conclusions:

• The EBST tool offers a robust and effective approach for identifying microRNA biomarkers for early ovarian cancer detection.
• The identified microRNAs hold significant promise for developing novel diagnostic strategies for ovarian cancer.
• The EBST tool and its associated MATLAB code are publicly available to facilitate further research.