FitHiC

FitHiC assigns statistical confidence to intra-chromosomal contact maps from Hi-C data to identify significant mid-range chromatin interactions (≈50 kb–10 Mb) linking regulatory elements and promoters.

Key Features:

• Statistical Confidence Estimation: Assigns statistical confidence estimates for mid-range (≈50 kb–10 Mb) intra-chromosomal Hi-C contacts to prioritize significant interactions.
• Joint Modeling Approach: Jointly models random polymer looping effects and technical biases to construct empirical null models of contact probability without relying on parametric distribution assumptions.
• Correction for Binning Artifacts: Corrects for binning artifacts in Hi-C contact maps to reduce spurious contacts introduced by binning.
• Improved Statistical Power: Provides improved statistical power relative to previous methods for detecting significant genomic interactions.

Scientific Applications:

• Linking Regulatory Elements and Promoters: High-confidence contacts preferentially connect expressed gene promoters with active enhancers as indicated by chromatin signatures in human embryonic stem cells (ESCs).
• Validation of Interactions: Captures 77% of RNA polymerase II-mediated enhancer-promoter interactions identified by ChIA-PET in mouse ESCs and confirms previously validated, cell line-specific interactions in mouse cortex cells.
• Insulator and Heterochromatin Analysis: Reveals insulators and heterochromatin regions serve as hubs for high-confidence contacts while promoters and strong enhancers are involved in fewer such contacts.
• Pluripotency Factor Binding Peaks: Binding peaks of NANOG and POU5F1 are highly enriched in high-confidence contacts within human ESCs.
• Replication Timing and Topological Domains: Loci linked by high-confidence contacts exhibit similar replication timing across human and mouse ESCs and preferentially lie within boundaries of topological domains.

Methodology:

Jointly model random polymer looping effects and technical biases to build empirical null models of contact probability without distributional assumptions, correct for binning artifacts, and assign statistical confidence to mid-range intra-chromosomal Hi-C contacts.