IGSA: Individual Gene Sets Analysis, including Enrichment and Clustering

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Traditional gene set analysis methods overlook sample expression heterogeneity, potentially missing key gene sets and failing to reflect disease severity.
• Existing approaches struggle to provide individual-level gene set expression profiles or link them to disease progression.

Purpose of the Study:

• To develop an application software, IGSA, for enhanced gene set enrichment and sample clustering.
• To address limitations in traditional gene set analysis by incorporating sample heterogeneity and disease severity.

Main Methods:

• IGSA calculates gene set expression scores per sample and employs an accumulating clustering strategy to group samples by disease progression.
• Performance was evaluated using gastric, pancreatic, and ovarian cancer datasets, with comparisons to DAVID, GSEA, SPIA, and ssGSEA for KEGG pathway enrichment.

Main Results:

• IGSA demonstrates superior sensitivity and specificity in identifying significant pathways compared to other methods.
• The software effectively indicates pathway changes related to disease severity and provides individual gene set profiles.
• IGSA's clustering approach successfully orders samples according to disease progression.

Conclusions:

• IGSA offers a powerful tool for gene set enrichment analysis and sample clustering, improving upon traditional methods.
• The software accurately reflects disease severity and provides valuable individual gene set expression profiles.
• IGSA enhances the understanding of biological pathways in cancer research.