deDoc2

deDoc2 identifies and characterizes hierarchies and dynamics of topologically associating domain-like domains (TLDs) in single-cell chromatin contact maps to enable analysis of chromatin architecture in relation to cellular identity and the cell cycle.

Key Features:

• Dynamic programming approach: Uses dynamic programming to detect genome partitions that minimize structure entropy across whole-genome and local contact matrices.
• Hierarchical TLD detection in single cells: Reported as one of two tools capable of identifying hierarchical TLD structures in single cells and demonstrated to surpass state-of-the-art alternatives in accuracy and reliability.
• Dynamic hierarchy analysis: Reveals that TLD hierarchies are highly dynamic across the cell cycle and vary among different human brain cortex cells.
• Association with cellular identity and function: Enables correlation analyses between TLD hierarchy and cellular identity and function to investigate potential regulatory roles of chromatin organization.

Scientific Applications:

• Single-cell TLD hierarchy mapping: Identification and characterization of hierarchical TLDs at single-cell resolution.
• Cell-cycle chromatin dynamics: Analysis of how TLD hierarchies change through the cell cycle.
• Cell-type specific chromatin organization: Comparison of TLD hierarchy variability among cell types, including human brain cortex cells.
• Regulatory mechanism inference: Exploration of links between chromatin hierarchical structure and cellular identity or function to inform gene-regulatory hypotheses.

Methodology:

Applies a dynamic programming algorithm to identify genome partitions that achieve minimal structure entropy using whole-genome and local contact matrices.