Hi-C Contact Map

Self-interacting genomic regions appearing as dense triangles along the diagonal of the contact map. Within a TAD, loci interact with each other far more frequently than with regions outside the domain boundary.

• Size range: 100 kb to 1 Mb in mammals
• Boundary identification: Sharp transitions in the Insulation Score
• Functional role: Constrain enhancer-promoter interactions to specific regulatory neighborhoods
• Disruption consequence: TAD boundary deletions can cause enhancer adoption and gene misexpression, linked to congenital disorders and cancers

Focal high-intensity points off the main diagonal representing physical contacts between two distal loci, typically anchored by CTCF and cohesin. These punctate signals indicate stable, recurrent chromatin loops.

• Loop anchors: Convergent CTCF motif orientation is a strong predictor
• Detection method: Peak-calling algorithms like HiCCUPS identify statistically enriched pixel pairs
• Functional significance: Bring enhancers into proximity with gene promoters over distances up to 1-2 Mb
• Dynamic nature: Loops can be cell-type specific and remodeled during differentiation

A plaid pattern visible at megabase resolution, reflecting the spatial segregation of chromatin into two mutually exclusive compartments.

• Compartment A: Gene-rich, transcriptionally active, open chromatin (euchromatin). High GC content.
• Compartment B: Gene-poor, transcriptionally silent, closed chromatin (heterochromatin). Associated with the nuclear lamina.
• Matrix signature: Positive correlation values (A-A, B-B) form the plaid; negative values (A-B) indicate spatial separation.
• B-to-A switching: Observed during cell differentiation and oncogene activation

The power-law relationship between contact probability and genomic distance. The contact frequency P(s) scales as s⁻¹ for intermediate distances, reflecting the fractal globule polymer model.

• Plot: Average contact probability vs. linear separation on a log-log scale
• Deviation: Sharp peaks above the baseline indicate loops; plateaus indicate TADs
• Physical interpretation: The exponent reveals the degree of chromatin compaction and folding
• Cross-cell comparison: Different cell types and species exhibit distinct P(s) curves, reflecting global differences in genome architecture

Genomic loci where the Insulation Score reaches a local minimum, indicating a sharp reduction in contacts across the boundary. These define the edges of TADs and are enriched for architectural proteins.

• Calculation: Sliding window sum of contacts crossing each bin, normalized by local background
• Protein enrichment: Strongly bound by CTCF, often with cohesin co-occupancy
• Directionality Index: Complementary metric showing upstream vs. downstream interaction bias
• Evolutionary conservation: TAD boundaries are more conserved than TAD interiors, suggesting functional constraint

Regions where multiple loci converge to form high-order interaction clusters, visible as vertical or horizontal stripes in the contact map. These often represent super-enhancers or Polycomb-repressed domains.

• Stripe pattern: A line of enriched contacts extending from a single anchor point
• Loop extrusion signature: Stripes indicate a unidirectional extrusion process stalled at a boundary
• Super-enhancer hubs: Clusters of enhancers that physically coalesce to drive high-level expression of cell-identity genes
• Polycomb bodies: Repressive hubs formed by H3K27me3-marked regions, silencing developmental genes

A high-resolution variant of chromosome conformation capture that uses micrococcal nuclease to fragment chromatin to the nucleosome level. Unlike standard Hi-C, which uses restriction enzymes and produces kilobase-scale fragments, Micro-C achieves mononucleosome resolution, enabling the detection of fine-scale interactions such as enhancer-promoter loops and transcription factor binding events that are invisible to coarser methods.

A scalable, chunked data format for storing very large, sparse genomic interaction matrices. Built on HDF5, Cooler enables efficient random access and out-of-core computation, allowing deep learning pipelines to query specific genomic regions without loading the entire multi-gigabyte contact map into memory. It supports multi-resolution storage via a hierarchical binning scheme.

A reproducibility metric specifically designed for Hi-C data that measures the similarity between two contact maps while accounting for the distance-dependent signal. Unlike Pearson correlation, which is dominated by the strong diagonal decay, SCC computes correlation within each genomic distance stratum independently and then aggregates, providing a fair benchmark for sequence-to-contact prediction accuracy.

The systematic correction of systematic biases in Hi-C contact matrices prior to analysis. Key biases include:

• GC content of restriction fragments
• Mappability of sequencing reads
• Restriction fragment length Methods like iterative correction and eigenvector decomposition (ICE) and matrix balancing transform raw counts into normalized interaction frequencies, ensuring that observed contacts reflect true biological proximity rather than technical artifacts.

The use of fluorescence in situ hybridization to experimentally measure the physical distance between specific genomic loci in fixed cells. By labeling two or more loci with distinct fluorophores and imaging their spatial separation, DNA FISH provides a gold-standard orthogonal validation for computationally predicted 3D genome structures, directly confirming whether predicted loops and contacts exist in physical space.

The prediction or analysis of 3D genome organization separately for maternal and paternal chromosomes. This haplotype-resolved approach reveals how genetic variation—such as single nucleotide polymorphisms or structural variants—differentially influences chromatin structure on each homologous chromosome. It requires phased sequencing data and specialized computational pipelines to partition Hi-C reads by parental origin.