TAD-Lactuca

TAD-Lactuca predicts topologically associating domain (TAD) boundaries by integrating histone modification signals and primary DNA sequence information to infer chromatin architecture.

Key Features:

• Integration of Epigenetic and Sequence Data: Utilizes histone modification signals and primary DNA sequences to inform TAD boundary prediction.
• Machine Learning Algorithms: Implements Random Forests (RF) and Multilayer Perceptrons (MLP) in Python for predictive modeling.
• Stability Across Resolutions: Maintains stable performance across different genomic resolutions and diverse datasets.
• High Predictive Accuracy: Achieves high predictive accuracy and surpasses state-of-the-art methods when sequence patterns are incorporated.

Scientific Applications:

• Understanding Chromatin Architecture: Supports analysis of 3D chromatin structure by predicting TAD boundaries relevant to cell differentiation and development.
• Gene Regulation Insights: Identifies sequence motifs enriched at TAD boundaries, providing insights into gene regulation mechanisms.
• Conservation Across Organisms: Supports comparative studies given conservation of TAD structures across different organs, relevant to evolutionary biology and developmental genetics.

Methodology:

Analyzes contextual information from epigenetic modifications and primary DNA sequences, applies Random Forests (RF) and Multilayer Perceptrons (MLP) implemented in Python, and identifies enrichment of transcription factor binding motifs at predicted TAD boundaries beyond the CCCTC-binding factor (CTCF) motif.