PB-DiffHiC: a statistical framework for detecting differential chromatin interactions from high resolution...

Area of Science:

• Genomics and Bioinformatics
• Molecular Biology
• Computational Biology

Background:

• Single-cell Hi-C (scHi-C) enables detailed analysis of chromatin interactions for understanding genome function.
• Analyzing high-resolution (10 Kb) scHi-C data is challenging due to data sparsity.
• Current methods like imputation or pseudo-bulk analysis have limitations.

Purpose of the Study:

• To develop a robust statistical framework for high-resolution differential chromatin interaction analysis using raw pseudo-bulk scHi-C data.
• To address the challenges posed by data sparsity in scHi-C analysis.
• To provide a computationally efficient and reliable alternative to existing methods.

Main Methods:

• Developed PB-DiffHiC, a parametric statistical framework for analyzing raw pseudo-bulk Hi-C data.
• Incorporated Gaussian convolution, short-range interaction stability, and Poisson modeling for normalization and testing.
• Benchmarked against existing methods using cell-type-specific chromatin loops and matched bulk Hi-C data.

Main Results:

• PB-DiffHiC demonstrated higher precision in identifying differential chromatin loops compared to alternative methods.
• Analysis of pseudo-bulk and matched bulk Hi-C data showed strong concordance, validating PB-DiffHiC's reliability.
• Successfully identified Kcnq5-associated differential interactions in a case study, aligning with other methods.

Conclusions:

• PB-DiffHiC is a statistically sound and robust method for high-resolution differential chromatin interaction analysis.
• The framework effectively handles raw pseudo-bulk Hi-C data, overcoming sparsity limitations.
• PB-DiffHiC offers a valuable tool for researchers studying genome structure-function relationships.