genomic-loop-call-interpretation

genomic-loop-call-interpretation

Summary

Interpret and annotate chromatin loop features detected on pre-generated Hi-C contact maps using Juicer's command-line feature annotation tools. This skill enables systematic extraction of loop coordinates and domain boundaries from kilobase-resolution Hi-C data, yielding structured feature annotations for downstream genomic analysis.

When to use

You have a pre-generated .hic contact map file and need to identify and annotate chromatin loops or topologically associating domains (TADs) at high resolution. Apply this skill when your research question requires extracting loop coordinates, domain boundaries, or other structural features from Hi-C maps rather than generating the contact maps themselves from raw sequencing data.

When NOT to use

• Input is raw FASTQ sequencing reads—use the Juicer pipeline stage for Hi-C map generation first
• You need to generate Hi-C maps from sequencing data—this skill assumes maps already exist; use the Juicer processing pipeline instead
• Your .hic file is incomplete or corrupted—verify file integrity before annotation

Inputs

• .hic contact map file (binary Hi-C matrix format)
• genome reference identifier (e.g., hg19, mm10)
• feature annotation parameters (loop resolution threshold, domain size cutoffs)

Outputs

• loop coordinate list (anchor pair positions in genomic coordinates)
• domain boundary file (TAD edge positions)
• annotated feature file (tool-specific format: loop anchors or domain boundaries with significance scores)

How to apply

Load the pre-generated .hic contact map file into the Juicer command-line tools suite. Select the appropriate feature annotation algorithm based on your target features—loop-calling tools (e.g., HiCCUPS) for interaction anchors or domain-detection algorithms for TAD boundaries. Execute the selected tool with appropriate parameters (cluster-size thresholds, significance cutoffs, GPU acceleration if available for HiCCUPS). The tool outputs annotated feature coordinates in a structured format (e.g., loop anchor positions, domain boundaries with genomic coordinates). Validate output by checking that coordinates fall within the expected genomic range and that feature lists are non-empty.