Arbor phylogenetics methods exposed through Docker.
List the methods available through Docker:
docker run kitware/arbordock
Run a method through Docker:
docker run <docker_options> kitware/arbordock <arg1> ...
In order to send data files to Docker, mount a volume and use the mounted
volume prefix for the input and output paths. For example, if anolis.phy
and anolis.csv are in your current directory (these are available in
the Git repository), the following will produce svl-model.csv in the current
directory:
docker run -v $PWD:/data kitware/arbordock /data/anolis.phy /data/anolis.csv lambda SVL /data/svl-model.csv
To build the Docker image locally:
git clone https://github.com/arborworkflows/arbordock.git
cd arbordock
docker build -t arbordock .
Now use arbordock instead of kitware/arbordock in the commands
above to use your local version.
- Install Girder.
- Enable the "Item tasks" plugin.
- Start a
girder_workerCelery worker and point it and the Worker plugin settings to use the same task queue. - From any folder you have write access to (you must be admin), select "Add tasks" from the actions menu.
- Enter "kitware/arbordock" as the image name and click Run. Wait for the task to complete, which could take a few minutes.
- Upload
anolis.phy(tree) andanolis.csv(character matrix) into Girder. - Go to "Phylogenetic signal" from the Tasks page in the left sidebar.
- Fill in the task parameters (enter a column name from the character matrix such as "SVL" for the column parameter) and click Run.
- Click "View output item" under "Outputs" on the task page to see the resulting model fit parameters.