Functional Analysis Tools For Metabolites: A Review (Docker Images)

This repository hosts Dockerfiles for all non-web-based methods covered in the review on functional analysis of metabolomics data.

QUICK START

Users should make sure they have Docker Desktop installed already.

Step 1: Clone this repository and save it to a folder on your local machine

# If you are users with a Linux/MAC computer, open your terminal
# If you are users with a Windows computer, open your Windows Powershell
cd Desktop                              
git clone https://github.com/tinnlab/metabolite-pathway-review-docker
cd metabolite-pathway-review-docker

Step 2: To build the docker image for a method of your interest, please run the following command:

# build base image first
docker-compose build base

# build a certain method image
docker-compose build <method-name>

# run the method image
docker-compose run --rm <method-name> bash

NOTE:

• Users are free to replace <method-name> with one of the following tool names: fella, lilikoi, metabolyzer, metax, mummichog, metaboanalyst, papi, puma, sspa, mbpls, ogpls, imsea, or dcimsea.

• We DO NOT recommend users to build all method images at once using docker-compose build as it can deplete your memory. Instead, build one tool at a time and remove it (docker rmi <method-name>) before proceeding to the next tool.

• For methods that require MATLAB to run (MB-PLS, iMSEA, and ogPLS), please visit the Matlab license center to obtain the license.lic file in advance. The hostid for this license should match the hostid of the host machine. The user of this license must be root. Once you had the license file, place it in the license sub-folder within the corresponding method's folder (i.e., MBPLS/license/, iMSEA/license/, or ogPLS/license/).