Visualize Data in HiGlass

This page provides instructions for visualizing DRAGEN secondary analysis outputs using HiGlass.

Cooler (.cooler) is a widely used file format for storing genomic contact matrices (e.g., Hi-C–style interactions). TruPath uses the cooler format to encode the frequency of reads aligned to pairs of genomic regions that were physically close on the flowcell, which forms a genome-wide contact map. Cooler is designed to efficiently store and query these large, sparse interaction matrices. The DRAGEN colocation cooler file can be found as <sample_name>/<sample_name>.colocation.cooler

HiGlass (https://higlass.io/), a community visualization tool for browsing genomic interaction data with smooth zooming, panning, and region-level inspection is recommended to explore colocation maps interactively.

---

Step 1 — Install HiGlass Docker

Use higlass-docker and follow instructions here

docker pull higlass/higlass-docker # Ensure that you have the latest.
docker run --detach \
           --publish 8899:80 \
           --volume  <local_path_to_mcooler>:/data \
           --name higlass-container_8899 \
           higlass/higlass-docker

Then, go to http://localhost:8899/app

Step 2 — Install cooler locally

You need to install cooler first before you can run commands like cooler zoomify.

You can install via pip:

pip install cooler

Or via conda (often preferred for scientific Python setups):

conda install -c conda-forge -c bioconda cooler

Step 3 — Convert .cooler to multi-resolution .mcooler

Once installed, you can generate an .mcool file using cooler zoomify, which recursively coarsens the matrix into multiple resolutions:

cooler zoomify <sample_name>.colocation.cooler

This produces <sample_name>.colocation.mcooler with zoom levels typically increasing by factors of 2 (so HiGlass remains responsive while zooming).

Step 4 — Visualize the .mcooler file with HiGlass

Next, Ingest mcooler from Step 2 into higlass:

higlass-manage ingest \
    --filetype cooler \
    --datatype matrix --assembly hg38 \
    --name <sample_name> --project-name <project_name> \
    <sample_name>.colocation.mcooler

In your HiGlass URL (e.g. http://localhost:8888/app), click the "+" on the top right > "Center"\

Next, select your sample colocation by <project_name>/<sample_name>, as you supplied to the higlass-manage ingest command. Here, my data is in cooler-test/test-chr14.

There are also several annotation tracks that are useful, i.e. you can add tracks on top and left for chromosome and genes. You can also turn on "Toggle Position Search Box" and you can navigate to your region of interest.

t(14,21)

HiGlass also has several add-ons to support VCF, bed, and bedpe file formats. Please feel free to add those tracks on the X/Y axis and explore the proximity reads from Constellation.