Deciphering the association between gene function and spatial gene-gene interactions in 3D human genome conformation

Area of Science:

• Genomics
• Bioinformatics
• Systems Biology

Background:

• Traditional gene function prediction methods utilize sequence identity, gene expression, and co-evolution.
• The role of three-dimensional (3D) genome conformation in gene function analysis has been largely unexplored due to limited data.
• Recent advancements in chromosome conformation capture techniques provide novel data for exploring genome architecture.

Purpose of the Study:

• To investigate the relevance of genome-wide spatial gene-gene interactions in predicting gene function.
• To develop and evaluate a novel gene function prediction method leveraging 3D genome conformation data.
• To assess the complementarity of spatial interaction data with genomic distance and sequence identity.

Main Methods:

• Construction of genome-wide spatial gene-gene interaction networks from Hi-C chromosomal contact data in human B-cells.
• Calculation of gene function similarity using G-SESAME and Fast-SemSim algorithms.
• Utilizing Gene Ontology (GO) statistics derived from interaction networks for function prediction.

Main Results:

• Genes with strong spatial interactions exhibit significantly higher functional similarity across biological processes, molecular functions, and cellular components.
• Interacted genes with high function similarity show stronger interactions, shorter genomic distances, and higher sequence identity.
• Integrating spatial interaction data with genomic distance or sequence identity substantially improves gene function similarity prediction.

Conclusions:

• The spatial conformation of the human genome is a significant factor influencing gene function similarity.
• 3D genome organization provides valuable, complementary information for accurate gene function prediction.
• This study highlights the utility of Hi-C data and network analysis for advancing our understanding of gene function.