cLoops

cLoops identifies chromatin loops from 3D genomic sequencing data such as ChIA-PET, Hi-C, HiChIP, and Trac-looping by clustering paired-end tags to call significant looping interactions.

Key Features:

• Data-Type Independence: Uses the cDBSCAN clustering algorithm to directly analyze paired-end tags (PETs) rather than relying on resolution-limited contact matrices or pre-called peaks.
• Statistical Significance Estimation: Classifies PETs into self-ligation and inter-ligation clusters and assesses significance of inter-ligation clusters using a permuted local background.
• Reduced Bias and Enhanced Enrichment: Produces loops with reduced distance-dependent bias and higher enrichment relative to local regions compared to other loop callers.
• Versatility and Efficiency: Handles both sharp and broad peak data and operates with modest computational resources.

Scientific Applications:

• Fine-scale loop detection: Detection of chromatin loops formed by interactions between regulatory elements spanning hundreds of kilobases.
• Cross-platform loop calling: Calling and comparing loops across ChIA-PET, Hi-C, HiChIP, and Trac-looping datasets.
• Genome regulation studies: Enabling analyses of gene regulation, epigenetic modifications, and genome architecture through accurate loop identification.

Methodology:

Applies cDBSCAN clustering to PETs, separates self-ligation and inter-ligation clusters, and evaluates inter-ligation cluster significance using a permuted local background.