HiCNAtra

HiCNAtra corrects Hi-C and 3C-seq contact maps from cancer cell lines for copy number variations and other systematic biases to enable more accurate chromatin conformation analysis.

Key Features:

• Read Depth Signal Computation: Computes read depth (RD) signals from Hi-C or 3C-seq reads using "entire restriction fragment" counting to generate high-resolution RD and CNV profiles.
• CNV Detection (CNAtra): Calls copy number events from the computed RD signal using the CNAtra methodology to identify CNVs that affect interaction frequencies.
• Bias Correction in Contact Maps: Integrates CNV information with other systematic biases and applies a Poisson regression model to correct chromatin interaction matrices toward euploid-equivalent conditions.
• Input and Platform: Processes HDF5 files produced by hiclib after iterative-mapping from Hi-C/3C-seq data and is implemented in MATLAB.
• Modular Architecture: Implements separate RD Calculator, CNV Caller, and Contact Map Corrector modules corresponding to RD computation, CNAtra-based CNV calling, and Poisson-regression-based matrix correction.

Scientific Applications:

• Cancer genomics and epigenetics: Produces corrected Hi-C contact maps for improved modeling and interpretation of genome-wide chromatin conformation in cancer cell lines, aiding studies of spatial chromatin organization and its implications in oncogenesis.

Methodology:

Processes HDF5 files from hiclib after iterative-mapping; computes RD signals via entire restriction fragment counting (RD Calculator); detects CNVs using the CNAtra CNV Caller on the RD signal; integrates CNV and other biases and corrects interaction matrices using a Poisson regression Contact Map Corrector.