hichub 1.1.0

Comprehensive Network Analysis for HiC

NOTE: For Paper Review, please follow the instruction below:

Latest updated on 02/22/2022,

Comprehensive Network Analysis for HiC


• Overview
• Documentation
• System Requirements
• Installation Guide
• Example of Running (Demo)
• License

Overview

This module is a Python package containing tool for network analysis of HiC data. It starts from HiC Interaction pairs, then generating network and clustering. Finally ranking all clusters by their interaction change.

System Requirements

Hardware Requirements

This package requires only a standard computer with enough RAM to support the in-memory operations.

Software Requirements

HicHub mainly depends on the Python scientific stack.

python >=3
pandas
numpy
pybedtools
python-igraph
scipy

Installation Guide

Quick install: python -m pip install git+https://github.com/WeiqunPengLab/hic_hub_test

Alternative method of installation: Recommend to use bioconda for installing. Create the environment from the environment_hichub.yml (Can be found in this repository) file:

conda env create -f environment_hichub.yml
python3 -m pip install hichub --user
python3 -m pip install numpy pandas pybedtools python-igraph scipy

https://bioconda.github.io/user/install.html

Example of Running (Demo)

After installation, type hichub in your command line will return the following UI:

welcome
The python env is: 3.7.10 | packaged by conda-forge | (default, Feb 19 2021, 16:07:37)
[GCC 9.3.0]
usage: hichub [-h] [-v] {test,diff,convert} ...
hichub -- A toolset to detect and analyze differential Hubs
positional arguments:
  {test,diff,convert}  sub-command help
    test               Output for test parser
    convert            Convert multi .hic to txt format --> Format should be: #chr bin1 bin2 Count
    diff               Parser for diff hubs
    find               Find gene in which hubs
    plot               Draw community plot to find gene in hubs
optional arguments:
  -h, --help           show this help message and exit
  -v, --version        show program's version number and exit

diff: Find hichubs for different cell conditions of the same specious.

usage: hichub diff [-h] -i <file> -f <str> -b <str> -r <int> [-d <float>] [-c <float>] [-p <float>] [-t <int>]

-i: Path to Input HiC file in txt format.
-f: Name of condition as foreground.
-b: Name of condition as background.
-r: Resolution if Hic .txt
-d: Foldchange to evaluate difference, default 1.0.
-c: Cut-off value to remove low abundance interaction pairs, default 10.0.
-p: Pvalue cutoff for output (diff hub), default 1e-5.
-t: Number of threads to run, default 1.

hichub diff --help 
Input Format: HiC Interaction in txt format. Example of test data can be found in ~/test_data
And the output can be found at working directory: 
(Demo output is: 6_1.0_DKO_na_WT_na_specific_regions.bed or 4_1.0_WT_na_DKO_na_specific_regions.bed)

convert: Convert (.hic) files to required input format

usage: hichub convert [-h] -i <file> [-n <str>] -f <str> -l <str> -r <int>

Exmaple of output: 
reg1    reg2    -log10(pvalue)
chr10:19570000-19850000 chr10:19570000-19850000 45.91186483627381
chr10:20860000-21060000 chr10:20860000-21060000 41.129022601906215
chr10:117030000-117140000       chr10:116870000-117010000       14.165
chr10:95130000-95290000 chr10:95130000-95290000 9.80623027538454
chr10:18970000-19160000 chr10:18970000-19160000 9.288099829570816

find: Find gene in which hub.

usage: hichub find [-h] -i <input file path> -f <str> -l <str> -n <gene name>
-i: Path to store all required input files.
-f: File names for different hubs, the output files for 'diff' call hubs. <file_name_1,file_name_2,...>
-l: The label for your input files. <condition_1, condition_2,...>
-n：The gene's name that you want to find in hubs. Examples: for human, LSR, SOX2, SERPINB8

plot: Plot community igraph-plot for the community that contain the gene you want to find in your hubs.

usage: hichub plot [-h] -i <input path for your .hic file> -c <input cluster files form 'diff'> -n <gene name> -f <str> -b <str>
-i: Path to Input HiC file in txt format.
-f: Name of condition as foreground.
-b: Name of condition as background.
-n: Gene name that you want to plot.
-c: Path to Input cluster file calcuated from 'diff'.

Built With

Contributing

Please read (https:xx) for details on our code of conduct, and the process for submitting pull requests to us.

Versioning

Authors

• *Xiang Li

Contributors

• Shuang Yuan

License

#This project is licensed under the MIT License - see the LICENSE.md file for details

Acknowledgments