scHiCDiff

scHiCDiff detects differential chromatin interactions (DCIs) in single-cell Hi-C data to identify cell type- or condition-specific variations in chromatin structure.

Key Features:

• Nonparametric and parametric approaches: Implements both nonparametric statistical tests and parametric models to provide alternative analytical frameworks for DCI detection.
• Zero-Inflated Negative Binomial model: Includes a zero-inflated negative binomial (ZINB) model to handle overdispersion and excess zeros in single-cell Hi-C contact counts and improve detection of DCIs between conditions.
• Comprehensive evaluation: Methods have been evaluated using simulated and real single-cell Hi-C datasets to assess performance in DCI detection.
• Cell type-/state-specific analysis: Enables identification of cell type- or state-specific chromatin interaction changes at single-cell resolution.
• Implementation: Implemented in R.

Scientific Applications:

• Chromatin dynamics: Analysis of chromatin interaction dynamics at single-cell resolution to characterize structural variation across cells.
• Comparative studies: Comparison of chromatin interactions between cell types or conditions to identify differential contacts associated with cellular states.
• Regulatory and epigenetic investigation: Linking DCIs to gene expression, cellular differentiation, and broader epigenetic regulation relevant to health and disease.

Methodology:

Applies nonparametric statistical tests and parametric models, including a zero-inflated negative binomial model, with methods evaluated on simulated and real single-cell Hi-C datasets and implemented in R.