Expression-based segmentation of the Drosophila genome

Area of Science:

• Genomics
• Molecular Biology
• Epigenetics

Background:

• Gene order in eukaryotes is nonrandom and crucial for gene regulation.
• Adjacent genes often exhibit shared expression patterns across tissues.
• Previous studies relied on correlation analysis to identify gene clusters.

Purpose of the Study:

• To develop a novel, probability-model-based method for identifying coexpression segments.
• To analyze gene organization and epigenetic associations in Drosophila melanogaster.
• To investigate the relationship between coexpression segments and physical interaction domains.

Main Methods:

• Developed a novel probability model to identify coexpression segments.
• Applied the method to Drosophila melanogaster gene expression data.
• Analyzed publicly available epigenetic data for associations.

Main Results:

• Two-thirds of Drosophila genes are part of multigenic coexpression segments (housekeeping and tissue-restricted).
• Adjacent genes within segments are physically closer; tissue-restricted segments are enriched for testis expression.
• Coexpression segmentation correlates with Hi-C domains but segments are smaller; intersegment regions are enriched for insulator binding sites.

Conclusions:

• A novel, heuristic-free method identifies widespread coexpression of adjacent genes in Drosophila.
• Coexpression segments represent a regulatory organization level below physical interaction domains.
• This organization is influenced by insulator binding sites.