cMonkey2 - Python port of the cMonkey biclustering algorithm
This is the Python implementation of the cMonkey algorithm based on the original R implementation by David J. Reiss, Institute for Systems Biology.
A complete set of documentation for installation and running of cMonkey is on the project's Github Pages.
There are also developer and user discussion groups.
Please report all bugs or other issues using the issue tracker. Please direct any and all questions to either the developer or user discussion groups.
The recommended way is to install cmonkey2 through pip
pip install cmonkey2
This will install the tools cmonkey2 and cm2view into your python environment. Please note that you will have to install MEME manually from http://meme-suite.org/
The simplest way to run the tool (if all data available in RSAT and STRING):
$ cmonkey2 --organism <organism-code> <tab separated file of gene expressions>
To display available options:
bin/cmonkey2.sh --help
To run the example organism:
bin/cmonkey2.sh --organism hal --rsat_base_url http://networks.systemsbiology.net/rsat example_data/hal/halo_ratios5.tsv
Below are the instructions to use cmonkey2 directly in the source repository
PreCyte made a Docker image based on cmonkey2 available on their github account
https://github.com/PreCyte/cMonkey2-docker/
cMonkey2 has been tested and runs on all tested recent versions of Linux (including debian-based [Ubuntu, Mint, Debian] and RPM-based [CentOS, Fedora]) and recent versions of Mac OS X. Additional dependencies include:
- Developed and tested with Python 2.7.x and Python 3.x
- scipy >= 0.9.0
- numpy >= 1.6.0
- biopython >= 1.63
- BeautifulSoup >= 4
- R >= 2.14.1
- rpy2 >= 2.2.1
- MEME 4.3.0 or >= 4.8.1 (all version up to 5.3.3 tested)
- csh (for running MEME)
- pandas
- sqlalchemy and sqlalchemy-utils
- svgwrite
for the human setup, Weeder 1.4.2 is needed
for running the unit tests (optional):
- python-xmlrunner
for running the interactive monitoring and visualization web application (optional):
- CherryPy 3
- Jinja2
- python-routes
bin/run_tests.sh
In general, you should be able to run cmonkey2 on microbial gene expression ratios with
bin/cmonkey2.sh --organism <organism-code> <tab separated file of gene expressions>
The file can be either in your file system or a web URL.
After the program was started, a log file will be written in cmonkey.log. You can see all available options with
bin/cmonkey2.sh --help
There is a startup script for cMonkey to run the current integrated system
bin/cmonkey2.sh --organism hal example_data/hal/halo_ratios5.tsv
bin/cm2view.sh [--out [output directory]]
bin/cmonkey2.sh --organism hal --rsat_organism Halobacterium_NRC_1_uid57769 --rsat_base_url http://pedagogix-tagc.univ-mrs.fr/rsat --rsat_features gene --nooperons --use_BSCM example_data/hal/halo_ratios5.tsv
To run cMonkey on human data, run the following code with your own <ratios.tsv> file
bin/cmonkey2.sh --organism hsa --string <stringFile> --rsat_organism Homo_sapiens_GRCh37 --rsat_URL http://rsat.sb-roscoff.fr/ --rsat_features protein_coding --nooperons <ratios.tsv>
Running cMonkey on Human data is somewhat difficult because neither the string database nor the RSAT database has human data cleanly entered. Here are the steps for a sucessful python cMonkey run on human
- Make a gene interaction file. The example data file mentioned above was generated from Biogrid around 10/6/14.
- Find an RSAT mirror that has .raw chromose files and feature files. In the above example, we use Homo_sapiens_ensembl_74_GRCh37 from the main RSAT database. To annotate these we use 'protein_coding.tab' and 'protein_coding_names.tab'. In principal, other annotation files such as 'processed_transcript' would work just as well.
- Adjust the upstream region searched, and perhaps modify the code to search for know TF and miRNA motifs rather than de-novo motifs. NOTE: Modiyfing the motif search step is non-trivial.
The distribution is built using setuptools and wheel format
- setup.py contains all information needed to build the distribution increase the version number before making a distribution
- record user-relevant changes in CHANGELOG.rst
python3 setup.py sdist bdist_wheel
twine upload -r pypi dist/cmonkey2-*