The COnstraint-Based Reconstruction and Analysis Toolbox

2 years after

The COBRA Toolbox - COnstraint-Based Reconstruction and Analysis Toolbox

MATLAB R2016b MATLAB R2015b MATLAB R2014b Code Coverage Code Grade
Build Status Build Status Build Status codecov Code grade


  1. Clone this repository

    git clone cobratoolbox
  2. Initialize submodules

    For convenience, we provide the SBMLToolbox 4.1.0, and glpk_mex in external/toolboxes, libSBML-5.13.0-matlab in io/utilities. Binaries for these librairies are provided in a submodule for Mac OS X 10.6 or later (64-bit), GNU/Linux Ubuntu 10.0 (64-bit), and Microsoft Windows 7 (64-bit). For unsupported OS, please refer to their respective building instructions (glpk_mex, libSBML).

    git submodule init
  3. Update submodules

    git submodule update
  4. From MATLAB, run


    NOTE: If you do not have gurobi_mex and tomlab installed on your machine, you will get some warnings and some errors. The COBRAToolbox will try to use glpk if it cannot find gurobi for LP / MILP. To solve any NLP problems you need tomlab_snopt.

  5. In order to test your installation, run


    to see what functions will work with your current configuration. It might be that some functions may not work unless you have tomlab with snopt installed.


All tutorials are included in the folder tutorials. More tutorial are currently being prepared.


The documentation is available on As this version is in development, you may find the legacy version of the documentation here.

Frequently Asked Questions (FAQ)

Before opening a new issue, please read first the section of Frequently Asked Questions FAQ.

How to contribute

How to cite The COBRA Toolbox

When citing The COBRA Toolbox, it is important to cite the original paper where an algorithm was first reported, as well as its implementation in The COBRA Toolbox. This is important, because the objective of The COBRA Toolbox is to amalgamate and integrate the functionality of a wide range of COBRA algorithms and this will be undermined if contributors of new algorithms do not get their fair share of citations. The following is one example how to approach this within the methods section of a paper (not the supplemental material please):

To generate a context-specific model the FASTCORE algorithm [1], implemented in The COBRA Toolbox [2], was employed.

[1] = Vlassis N, Pacheco MP, Sauter T (2014) Fast Reconstruction of Compact Context-Specific Metabolic Network Models. PLoS Comput Biol 10(1): e1003424.

[2] = Schellenberger J, Que R, Fleming RMT, Thiele I, Orth JD, Feist AM, Zielinski DC, Bordbar A, Lewis NE, Rahmanian S, Kang J, Hyduke DR, Palsson BØ. 2011 Quantitative prediction of cellular metabolism with constraint-based models: The COBRA Toolbox v2.0. Nature Protocols 6:1290-1307.


Read more on the compatibility with SBML-FBCv2

Related Repositories



High-level, high-performance, constraint-based reconstruction and analysis in Ju ...



MATLAB Development Tools ...