scHiCcompare

scHiCcompare implements differential chromatin interaction analysis between two single-cell Hi-C data groups through imputation, normalization, and statistical testing.

Key Features:

• Imputation by pooling: The package implements pooling techniques for imputation of sparse single-cell Hi-C data, considering genomic distances.
• Normalization: Data are normalized using LOESS regression.
• Differential analysis: Differential chromatin interactions are detected via Gaussian Mixture Model (GMM) clustering.
• Pseudo-bulk matrix generation: The approach creates pseudo-bulk Hi-C matrices by summing condition-specific data.

Scientific Applications:

• Cancer research: The method enables detection of changes in three-dimensional genome structure that serve as key indicators of cancer.
• Developmental disorder studies: The approach identifies 3D structural changes associated with developmental disorders.
• Cell type-specific chromatin structure analysis: The workflow detects cell type-specific differences in chromatin structure and their associations with biological and epigenetic features.

Methodology:

scHiCcompare builds on existing techniques in the HiCcompare R package. It imputes sparse single-cell Hi-C data by considering genomic distances, creates pseudo-bulk Hi-C matrices by summing condition-specific data, normalizes data using LOESS regression, and detects differential chromatin interactions via Gaussian Mixture Model (GMM) clustering.