SpectralTAD

SpectralTAD identifies Topologically Associated Domains (TADs) from Hi-C contact matrices using a sliding window-based spectral clustering framework that detects hierarchical domain structures for understanding genome regulation.

Key Features:

• Hierarchical TAD detection: Detects TADs, sub-TADs, and higher-level hierarchical structures from Hi-C data.
• Sliding window spectral clustering: Uses a modified spectral clustering approach with a sliding window for rapid TAD detection.
• Automatic parameter selection: Does not require users to adjust parameters for operation.
• Multiple input formats: Works on a range of different formats of contact matrices.
• Controlled hierarchy levels: Gives users control over the number of hierarchical levels to be returned.
• Robustness to data variation: Robust to sequencing depth, resolution, and sparsity of Hi-C data.
• Output format: Returns a bed file of TAD coordinates.

Scientific Applications:

• 3D genome structure analysis: Identifies hierarchical TAD structures critical for understanding genome regulation and gene expression.
• Boundary mark enrichment analysis: Enables identification of TAD boundaries enriched in classical boundary marks across multiple hierarchy levels.
• Cross-cell-line and cross-tissue conservation studies: Supports analysis of boundary conservation patterns to distinguish stable versus dynamic regulatory elements.
• Comparative TAD caller evaluation: Provides benchmarked performance against state-of-the-art TAD callers in simulated and experimental settings.

Methodology:

SpectralTAD uses a sliding window-based spectral clustering framework that identifies gaps between consecutive eigenvectors for TAD boundary identification.