A Cyclic Permutation Approach to Removing Spatial Dependency between Clustered Gene Ontology Terms.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Traditional gene set enrichment analysis struggles with large genomic regions where gene proximity can confound results.
• Spatial dependency in genomic data can lead to false positives, mistaking physical clustering for functional enrichment.

Purpose of the Study:

• To introduce Spatial Adjusted Gene Ontology (SAGO), a novel method to address spatial bias in gene set enrichment analysis.
• To differentiate genuine biological associations from those arising from spatial proximity in genomic domains.

Main Methods:

• Developed a cyclic permutation-based approach (SAGO) that integrates gene spatial arrangement.
• Applied SAGO to diverse large genomic datasets, including replication timing domains, epigenetic marks (H3K9me3, H3K27me3), HiC compartments, and lamina-associated domains (LADs).

Main Results:

• SAGO successfully separated spatial proximity effects from true biological enrichments across various genomic contexts.
• Application to prostate cancer copy number alteration (CNA) domains significantly reduced spurious GO term enrichments.
• The analysis identified biologically relevant gene sets linked to oncogenic processes, free from spatial bias.

Conclusions:

• SAGO provides a robust solution for accurate gene set enrichment analysis in the presence of spatial genomic dependencies.
• The method enhances the identification of functionally significant gene sets, particularly in cancer genomics and large-scale domain analyses.