Identifying Epigenetic Biomarkers using Maximal Relevance and Minimal Redundancy Based Feature Selection for...

Area of Science:

• Bioinformatics
• Genomics
• Epigenetics

Background:

• Epigenetic biomarker discovery is crucial in bioinformatics.
• Existing methods may lack efficiency in handling multi-omics datasets.
• Identifying statistically significant biomarkers requires robust feature selection.

Purpose of the Study:

• To develop a new framework for identifying statistically significant epigenetic biomarkers.
• To utilize maximal-relevance and minimal-redundancy criterion for gene selection in multi-omics data.
• To improve the positive predictive rate and reduce the false positive rate in biomarker identification.

Main Methods:

• Selected genes with both normal and non-normal distribution of expression and methylation values.
• Applied a gene-selection method for maximal-relevant, minimal-redundant genes.
• Utilized t-test for normally distributed genes and Limma package for non-normally distributed genes for statistical validation.
• Performed comparative analysis based on classification performance.

Main Results:

• Identified top-ranking significant gene-markers with biological validation.
• The proposed framework demonstrated improved positive predictive rates.
• Reduced false positive rates in epigenetic marker identification.
• Comparative analysis showed competitive classification performance.

Conclusions:

• The developed framework effectively identifies statistically significant epigenetic biomarkers from multi-omics data.
• The maximal-relevance, minimal-redundancy approach enhances biomarker discovery.
• The method offers a robust approach for epigenetic biomarker identification with improved accuracy.