Clodius

Quick start without docker

Install the clodius package:

pip install clodius

And use it aggregate a bigWig file:

clodius aggregate bigwig ~/Downloads/E116-DNase.fc.signal.bigwig

The output files can then be displayed using the higlass-docker container. For more information about viewing these types of files take a look at the higlass wiki.

Development

The recommended way to develop clodius is to use a conda environment and install clodius with develop mode:

python setup.py develop

Note that making changes to the clodius/fast.pyx cython module requires an explicit recompile step:

python setup.py build_ext --inplace

Testing

The unit tests for clodius can be run using nosetests:

nosetests tests

Individual unit tests can be specified by indicating the file and function they are defined in:

nosetests test/cli_test.py:test_clodius_aggregate_bedgraph

Quick start with Docker

If you don't have your own, get some sample data:

mkdir -p /tmp/clodius/input
mkdir -p /tmp/clodius/output
curl https://raw.githubusercontent.com/hms-dbmi/clodius/develop/test/sample_data/geneAnnotationsExonsUnions.short.bed \
  > /tmp/clodius/input/sample.short.bed 

Then install Docker, and pull and run the Clodius image:

docker stop clodius; 
docker rm clodius;

docker pull gehlenborglab/clodius # Ensure that you have the latest.

docker run -v /tmp/clodius/input:/tmp/ \
           gehlenborglab/clodius \
           clodius aggregate bigwig /tmp/file.bigwig
           
ls /tmp/clodius/input # Should contain the output file

If you already have a good location for your input and output files, reference that in the -v arguments above, instead of /tmp/clodius. The other scripts referenced below can be wrapped similarly.