Inspectro

Spectral decomposition for characterizing long-range interaction profiles in Hi-C maps.

Snakemake workflow for the unsupervised method presented in Spracklin, Abdennur et al., 2022.

@article{spracklin2022diverse,
  title={Diverse silent chromatin states modulate genome compartmentalization and loop extrusion barriers},
  author={Spracklin, George and Abdennur, Nezar and Imakaev, Maxim and Chowdhury, Neil and Pradhan, Sriharsa and Mirny, Leonid A and Dekker, Job},
  journal={Nature Structural \& Molecular Biology},
  pages={1--14},
  year={2022},
  publisher={Nature Publishing Group}
}

Steps

0. Set up your environment with conda or mamba:

conda env create -n spec -f environment.yml
conda activate spec

1. Place supplementary bigwig file information in config/track_metadata.tsv to include in graphical outputs. Must be tab-delimited, including header. Must have columns:

• Name: a display name
• ID: a unique identifier to use in the database (can be the same as Name)
• FileFormat: must be the string bigWig
• Path: a local path to the file

2. Edit config.yaml.

3. Run:

# Generate pre-aggregated supplementary bigwig tracks if provided.
$ snakemake make_track_db --force --cores all

# Run the pipeline
$ snakemake --cores all

Releases

Versions v0.1.* will always follow the decomposition and clustering method as described in the NSMB paper, incorporating bug fixes and maintenance.

Future versions of the pipeline (v0.2 and up) may include new additions and enhancements to the original method, which will be documented.