Selfish

Selfish detects differential chromatin interactions by comparing Hi-C contact maps using a structural self-similarity measure.

Key Features:

• Structural self-similarity measure: Computes a self-similarity measure on Hi-C contact maps to capture local and global structural patterns.
• Reproducibility assessment: Quantifies similarity between replicate Hi-C contact maps to assess experimental reproducibility.
• Differential chromatin interaction detection: Identifies variations in chromatin interactions between two contact maps to pinpoint differential genomic regions.
• Validation on simulated and real data: Demonstrated superior accuracy and robustness on simulated and real Hi-C datasets compared to existing methods.
• Python implementation: Provides a Python implementation of the algorithms for computational analysis of Hi-C contact maps.

Scientific Applications:

• Reproducibility analysis: Assess reproducibility of replicate Hi-C experiments by quantifying contact-map similarity.
• Differential interaction discovery: Detect differential chromatin interactions between conditions or samples to locate biologically significant regions.
• 3D genome and gene regulation studies: Compare contact maps to study dynamic genome architecture and its implications for gene regulation.

Methodology:

Compute a structural self-similarity measure on Hi-C contact maps and apply algorithms to compare pairs of contact maps for reproducibility assessment and differential interaction detection; validation used simulated and real Hi-C datasets.