Weighted-SAMGSR: combining significance analysis of microarray-gene set reduction algorithm with pathway...

Area of Science:

• Bioinformatics
• Computational Biology
• Systems Biology

Background:

• Pathway-based feature selection methods generally outperform gene-based methods in predictive accuracy and stability.
• Significance Analysis of Microarray-Gene Set Reduction algorithm (SAMGSR) is a pathway-based method that reduces pathways to core subsets.

Purpose of the Study:

• To propose a weighted version of SAMGSR that incorporates gene connectivity information.
• To improve the performance of pathway-based feature selection by considering pathway topology.

Main Methods:

• Developed a weighted SAMGSR algorithm using weights derived from gene connectivity.
• Combined these weights with test statistics for gene selection.
• Evaluated the algorithm using simulated and real-world data.

Main Results:

• The weighted SAMGSR algorithm demonstrated superior performance compared to the original SAMGSR.
• The inclusion of gene connectivity information significantly facilitated feature selection.

Conclusions:

• The proposed weighted SAMGSR effectively utilizes gene connectivity for enhanced feature selection.
• Incorporating topological information improves pathway-based gene selection accuracy and stability.