Multiple Criteria Optimization (MCO): A gene selection deterministic tool in RStudio

Area of Science:

• Bioinformatics
• Genomics
• Computational Biology

Background:

• Gene selection is crucial for understanding molecular mechanisms and therapeutic development.
• Reproducibility, objectivity, and repeatability are essential in biological and informatics analyses.
• Previous work proposed a multiple criteria optimization (MCO) algorithm for objective gene selection from microarray data.

Purpose of the Study:

• To develop an open-source R tool for gene selection using MCO.
• To enable both individual dataset analysis and meta-analysis of microarray data.
• To provide an affordable, repeatable, and objective method for detecting differentially expressed genes.

Main Methods:

• Development of an open-source R tool implementing the MCO algorithm.
• The tool supports individual analysis of datasets (2-3 performance measures) and meta-analysis (up to 5 datasets).
• Demonstration using four Parkinson's Disease (PD) microarray datasets for individual and meta-analysis.

Main Results:

• The MCO algorithm ensures objective and repeatable gene selection without user parameter manipulation.
• The R tool is portable, requires modest hardware, and is license-free.
• Analysis identified MMP9 and TUBB2A as potential PD genetic biomarkers due to their consistent presence across datasets.

Conclusions:

• The developed R tool offers an affordable, repeatable, and objective approach to gene selection from microarrays.
• The identified genes MMP9 and TUBB2A show potential as PD biomarkers, validated by literature.
• The tool is applicable to various array-based experiments, including microRNA and protein arrays.