Capturing cell type-specific chromatin compartment patterns by applying topic modeling to single-cell Hi-C data.

Area of Science:

• Genomics
• Molecular Biology
• Computational Biology

Background:

• Single-cell Hi-C (scHi-C) provides insights into 3D genome organization at the individual cell level.
• The inherent sparsity of scHi-C data presents significant analytical challenges, hindering the extraction of biological insights.
• Existing methods struggle to effectively analyze the high-dimensionality and sparsity of scHi-C datasets.

Purpose of the Study:

• To develop and apply a novel computational approach for analyzing sparse single-cell Hi-C data.
• To leverage topic modeling to identify latent structures and cell type-specific features within scHi-C data.
• To overcome the limitations posed by data sparsity in single-cell 3D genome organization studies.

Main Methods:

• Generation of nine single-cell combinatorial indexed Hi-C (sci-Hi-C) libraries from five distinct human cell lines, comprising over 19,000 cells.
• Application of unsupervised topic modeling algorithms to the generated sci-Hi-C datasets.
• Analysis of the discovered 'chromatin topics' for cell type-specific patterns and enrichment of genomic features.

Main Results:

• Topic modeling successfully identified distinct 'chromatin topics' from sparse sci-Hi-C data.
• These 'chromatin topics' effectively captured and differentiated between cell types based on their 3D genome organization.
• Enrichment analysis revealed specific compartment structures associated with locus pairs within the identified topics.

Conclusions:

• Topic modeling is a powerful and effective method for analyzing sparse single-cell Hi-C data.
• This approach enables the discovery of cell type-specific chromatin organization patterns.
• The findings offer new avenues for exploring the functional implications of 3D genome architecture in individual cells.