fagan2888 / gadget-pipeline

An analysis pipeline used to classify ET vs. PD tremor based on wrist-mounted accelerometer data.

Gadget PD Pipeline

A pipeline for differentiating and categorizing patients of Parkinsons and Essential Tremor using a mixture of genetic and motion tracking techniques.

Using this pipeline

This pipeline is built using metapipe and a variety of other tools. Install the dependencies as follows:

$ pip install -r requirements.txt

Once all the dependencies are installed, the next step is to execute the pipeline. If you wish to run it locally, then use the following command:

$ sh run_me.sh <sample_id> <API_USERNAME> <API_PASSWORD>

To run this pipeline on an external system, like a cluster, you'll need to edit the pipeline.mp file and fill in your api username, password, and sample id. When that is done, run your pipeline using these commands:

$ metapipe -j pbs -o pipeline.sh pipeline.mp 
$ qsub pipeline.sh

Project Organization

This project relies on a few external projects and a variety of shell scripts. The shell scripts can be found in the bin/ directory, and the related packages are in the requirements.txt.

The overall structure of the pipeline can be found in the pipeline.mp file. This file contains the metapipe-format pipeline design.