Installation

The intended use of this pipeline is through the use of a Singularity or Docker image. The image can be built using the Dockerfile found in the repository, or it can be pulled from Dockerhub as a singularity using the following command:

singularity pull docker://dcanumn/osprey:<version_num>

Where version_num denotes the specific version of the container. All available versions of the container can be found here.

After downloading the container, singularity is the only other dependency needed for processing. The full usage details can be seen under the usage section, but the basic command to run the container is as follows:

container_path=/path/to/container.sif
bids_dir=/path/to/bids
output_dir=/path/to/output
bibsnet_dir=/path/to/bibsnet
settings_file=/path/to/settings.json
singularity run -B $bids_dir:/bids \
 -B $output_dir:/output \
 -B $bibsnet_dir:/bibsnet \
 -B $settings_file:/settings_file/file.json \
 $container_path /bids /output participant /settings_file/file.json

Where “singularity run” is followed by specific commands for singularity. In this case it is a series of “bind” commands that will give singularity access to the necessary directories for processing. Of note, you will need to be sure that singularity is given access to the BIDS directory, the output directory where you want pipeline outputs to be stored, and the JSON file that will configure how processing is run.

The singularity specific arguments are then followed by the path to the container. Finally, the user specifies the arguments that are unique to the current application, such as input and output files, configuration files, and other processing settings.

Note: If you are not interested in interacting with this particular interface of OSPREY, which includes BIDS and containerized functionality, consider using OSPREY directly in Matlab following the instructions found on the OSPREY github.