Skip to main content

Hi-C Pipeline Overview

This section covers the analysis pipeline for Hi-C data, used to study genome-wide chromatin interactions and 3D genome organization.

What is Hi-C?

Hi-C is a genome-wide chromosome conformation capture technique that maps all chromatin interactions in the nucleus, as an all-to-all method. This displays the 3D organization of chromosomes including TADs (Topologically Associating Domains), compartments, and chromatin loops.

What is Tiled-C?

Tiled-C is a targeted version of Hi-C that focuses on specific regions of interest, by using a set of tiled capture probes across a wide region (~few Mbs) that then allows for higher resolution analysis of chromatin interactions in those regions. It is particularly useful for studying regulatory elements and their interactions with target genes. Despite been a targeted aproach the data is still analysed using the same Hi-C pipeline, but with a few modifications to account for the targeted nature of the data.

Pipeline Steps

  1. Quality Control - FastQC analysis of raw reads
  2. Pairing and Filtering - Pairing read ends, removing duplicates, filtering artifacts
  3. Alignment - Mapping paired-end reads to reference genome
  4. Assignment - Assigning reads to restriction fragments (at various resolutions)
  5. Valid Pair Generation - Creating interaction matrices
  6. Normalization - ICE normalization of contact matrices

Running the Hi-C pipeline

Once you’ve installed HiC-Pro for example into your mambe installation, you can run the pipeline using the following command:

/project/higgslab/your_username/mamba_installation/HiC-Pro/bin/HiC-Pro -i Input_fastq/ \
 -o HiC_Pro_output \
 -c HiC_Pro_scripts/config.txt

Where the config.txt is a configuration file that you can edit to specify the parameters for your analysis, including the reference genome, restriction enzyme, and other options. You can find an example config file in the HiC-Pro repository.

Typical Outputs

  • QC reports
  • Contact matrices at multiple resolutions

Things to consider when analysing Hi-C data

  • Resolution: The resolution of the contact maps depends on the sequencing depth and the size of the genome. Higher resolution requires more sequencing depth.
  • Normalization: Different normalization methods can be applied to account for biases in Hi-C data. ICE normalization is commonly used, but other methods may be more appropriate depending on the specific analysis.
  • Quality Control: It is important to assess the quality of the Hi-C data, including the percentage of valid pairs, duplication rates, and the distribution of interaction distances. High duplication rates or low valid pair percentages may indicate issues with the library preparation or sequencing.