Skip to content

Installation

This page describes the software requirements and installation steps for the TackleMMe analysis pipeline.

MATLAB version

The pipeline requires MATLAB R2024a or R2024b.

MATLAB R2025 is not compatible

The UMAP toolbox used by this pipeline relies on Java access to MATLAB figures, which is removed in MATLAB R2025. Use R2024b or earlier.

Check your MATLAB version in the command window:

ver

Required toolboxes

The following MathWorks toolboxes must be installed. Install them via Home > Add-Ons > Get Add-Ons in MATLAB, or through your institution's license portal.

Toolbox Required for Minimum version
Statistics and Machine Learning Toolbox UMAP, statistical analysis R2024a
MATLAB Report Generator Automated report generation R2024a
Cobra Toolbox Uses cobra model structure + functions 2024

Verify that all toolboxes are installed:

ver

The output should list both Statistics and Machine Learning Toolbox and MATLAB Report Generator.

UMAP toolbox installation

The pipeline uses the MATLAB UMAP implementation by Meehan et al. [File Exchange #71902].

Step 1 — Download

Download the toolbox from MATLAB File Exchange.

Alternatively, open it directly in MATLAB Online by clicking Open in MATLAB Online on the File Exchange page.

Step 2 — Add to MATLAB path

Unzip the downloaded archive to a permanent location (e.g. C:\Tools\umap), then add it to your MATLAB path:

addpath(genpath('C:\Tools\umap'));
savepath;

Save the path

Run savepath after addpath so the UMAP toolbox is available in future MATLAB sessions without re-adding it manually.

Step 3 — Verify installation

% Should run without errors and produce a 2D scatter plot
X = rand(100, 10);
Y = run_umap(X, 'n_components', 2);
scatter(Y(:,1), Y(:,2));

Dependencies

The UMAP toolbox requires the Statistics and Machine Learning Toolbox and the Financial Toolbox. Check with ver if you encounter missing function errors.

Report generation

The pipeline generates automated analysis reports using the MATLAB Report Generator toolbox. Reports are produced by the writeAnalysisReport function.

How it works

The report generator uses the Report API and Document Object Model (DOM) API to programmatically build documents from your analysis results. The pipeline captures tables, and text output from the single model analysis step and compiles them into a structured PDF report. The report file is saved to the working directory.

Verification checklist

After completing all installation steps, verify your setup with the following commands in the MATLAB command window:

% 1. Check MATLAB version (should be R2024a or R2024b)
ver('matlab')

% 2. Check required toolboxes
ver('stats')      % Statistics and Machine Learning Toolbox
ver('rptgen')     % MATLAB Report Generator

% 3. Check UMAP installation
which run_umap    % Should return the path to run_umap.m

% 4. Test UMAP
Y = run_umap(rand(50, 5), 'n_components', 2);
disp(size(Y))     % Should be [50, 2]

If all commands execute without errors, your environment is ready for the pipeline.

Installing the COBRA Toolbox

The COnstraint-Based Reconstruction and Analysis Toolbox is a MATLAB software to create & analyze constrained based metabolic models. All the models stored & worked with in TackleMMe are stored in COBRA format and functions of the COBRA Toolbox are used. Therefore the toolbox must be installed (see the github page as well as the documentation webpage).

Make sure to add the toolbox to your matlab path: 

addpath(genpath('path/to/toolbox/on/your/local/machine'));
savepath;

Building models with rFastcormics

rFASTCORMICS will be integrated into the next release of the COBRA Toolbox. In the meantime, if you want to build context-specific models using rFASTCORMICS before comparing them with TackleMMe, follow the installation instructions on the rFASTCORMICS GitHub page.

A supplementary rFastcormicsPipeline.m script is provided in the scr/ folder. It builds context-specific models and outputs a params structure that can be passed directly to createProject.m. For a usage example, see no1_modelBuildingAndProjectInit.m in the BRCAexample folder.

Install TackleMMe Pipeline

Clone the repository to your local machine. The zip download may not work due to the high size of some datasets. Therefore, we recommend to clone the repository instead.

git pull https://github.com/sysbiolux/analysisPipelineLVT.git