A multi-task graph-clustering approach for chromosome conformation capture data sets identifies conserved modules of...

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Chromosome conformation capture (3C) methods are vital for studying 3D genome architecture.
• Analyzing changes in genome structure across diverse cell types and species presents a significant challenge.

Purpose of the Study:

• To introduce Arboretum-Hi-C, a novel multi-task spectral clustering approach.
• To differentiate between conserved and context-specific aspects of genome architecture.

Main Methods:

• Arboretum-Hi-C employs multi-task spectral clustering.
• The method is applied to analyze genome architecture data.

Main Results:

• Arboretum-Hi-C yields more biologically consistent conservation patterns than standard clustering.
• Identified clusters are largely conserved and associated with high- or low-activity genomic signals.
• Most genomic regions show divergence between clusters with similar chromatin states, with exceptions in lamina-associated domains and open chromatin.

Conclusions:

• Arboretum-Hi-C effectively distinguishes commonalities and differences in 3D genome organization.
• The findings provide insights into the conservation and divergence of genomic architecture, particularly concerning chromatin states.