SCIA: A Novel Gene Set Analysis Applicable to Data With Different Characteristics.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Gene set analysis is crucial for functional enrichment and pathway analysis.
• Existing competitive methods amplify false discovery rates with correlated genes.
• Self-contained methods have data characteristic restrictions, necessitating a more robust approach.

Purpose of the Study:

• To develop a statistically rigorous gene set analysis method applicable to diverse datasets.
• To alleviate bias and improve comparability in functional enrichment analyses.
• To introduce a novel approach that combines self-contained and competitive testing strategies.

Main Methods:

• Proposed Self-contained and Competitive Incorporated Analysis (SCIA).
• Utilized a novel permutation strategy incorporating a priori biological networks.
• Selectively permuted gene labels with varying probabilities based on network information.

Main Results:

• SCIA demonstrated superior performance in simulation studies compared to GSEA, CAMERA, ROAST, and NES.
• Achieved better false discovery rate and sensitivity across various conditions.
• KEGG pathway analysis on lung cancer datasets identified more significant pathways with SCIA, many supported by literature.

Conclusions:

• SCIA offers a more reliable and comparable gene set analysis method.
• The novel permutation strategy effectively addresses limitations of existing methods.
• SCIA provides researchers with improved tools for functional enrichment and pathway discovery.