Accurate loop calling for 3D genomic data with cLoops

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• 3D genome mapping technologies identify interactions between distant regulatory elements.
• Existing loop-calling tools have limitations in data type compatibility, accuracy, and hardware requirements.

Purpose of the Study:

• To introduce cLoops, a novel tool for identifying 3D genome loops.
• To overcome the limitations of existing loop-calling methods.

Main Methods:

• cLoops utilizes the cDBSCAN clustering algorithm to directly analyze paired-end tags (PETs).
• It employs a permuted local background for statistical significance estimation.
• The tool is data-type independent, supporting various sequencing data like ChIA-PET, Hi-C, HiChIP, and Trac-looping.

Main Results:

• cLoops reliably identifies loops from both sharp and broad peak data.
• Loops detected by cLoops exhibit reduced distance-dependent bias and higher enrichment compared to existing tools.
• The tool demonstrates improved accuracy, versatility, flexibility, and efficiency with modest hardware needs.

Conclusions:

• cLoops offers an accurate, versatile, and efficient solution for 3D genome loop detection.
• The tool addresses key limitations of previous methods, making advanced genomic analysis more accessible.