GLPK

To install PyFBA we first need to install a linear solver. We use the GNU Linear Programming KitGLPK program as our linear solver, although you can use others. In addition, youwill need to install the Gnu Multiple Precision arithmetic library. Finally, we install PyGLPK, aPython wrapper around GLPK.

There are several Python language bindings to choose from. Each provides a differing level of abstraction. All are open source software. The Scripting plus. PuLP is an LP modeler written in python. PuLP can generate MPS or LP files and call GLPK, COIN CLP/CBC, CPLEX, and GUROBI to solve linear problems. PuLP is an LP modeler written in Python. PuLP can generate MPS or LP files and call GLPK 1, COIN CLP/CBC 2, CPLEX 3, and GUROBI 4 to solve linear problems.

Install GLPK on Linux

Installing GLPK on Ubuntu or other Debian systems

If you are on a Debian system you can install them with

apt-get install libglpk-dev glpk-utils glpk-doc libglpk36 libgmp10 libgmpxx4ldbl

(You maybe able to get away with less, but that is what I have installed).

Once you have that installed, you should be able to continue with the PyGLPK installation below

Installing GLPK on CentOS

If you are on a CentOS system, the easiest way to install this is to first install GMP, download the GLPK library fromGNU, and build it:

At this point, you need to add /usr/local/lib to your LD_LIBRARY_PATH and then continue with the PyGLPK installationbelow

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/lib python2.7 setup.py build python2.7 setup.py install

Installing GLPK on Mac OSX 10.10 Yosemite

Unfortunately, the easiest way is to build glpk from source

Download the latest version of GLPK from http://ftp.gnu.org/gnu/glpk/ and then open a Terminal and install GLPK:

Installing GLPK on Windows

You will need to install WinGLPK but their installation instructions are not veryclear. We will have detailed installation instructions shortly.

Install PyGLPK

You need to install PyGLPK. Make sure that you install this fork ofPyGLPK as the original has not worked for quite some time!

Install the ModelSEEDDatabase

We rely on the Model SEED to provide the biochemistry tables that we use (although wehave designed PyFBA with the potential to use biochemistry tables from other sources if they become publiclyavailable!). To get the latest ModelSEED database you need to clone their GIT repository to a destination on your harddrive. You then need to set the ModelSEEDDatabase environment variable as explainedbelow.

Python modules

PyFBA depends on a few different Python modules:

As noted above, you should install PyGLPK from GitHub.However, setup.py will try and do the right thing for you.

libSBML and lxml

One or two of the scripts (notably scripts/run_fba_sbml.py) require that you have libSBMLand lxml installed to read the SBML files.

setup.py will attempt to install these for you. If you wish to install them manually you should be able to do sowith pip install:

Beautiful Soup 4

XML parsing is a pain in the butt, and so we use Beautiful Soup 4to make life easy! setup.py should try and install this for you, but if you wish to do it manually, you shouldbe able to do so with pip intall:

You should be able to install PyFBA from PyPI using pip install:

If that does not work, you can clone the git hub repository and run setup.py manually:

If you do not have administrative (root) access to your machine you can also install the code in alocal directory:

PyFBA relies on the Model SEED Database, and we need to know where that is installed. We also provide some examplemedia files with the download and you can set an environment variable (PYFBA_MEDIA_DIR) that points to the location ofthose files if you want to include them for your models.

Set the following two environment variables:

For more information on setting the environment variables, see one of these sites:

• MacOS:
  • You can set them globally as described (here)[http://stackoverflow.com/questions/135688/setting-environment-variables-in-os-x/3756686#3756686]or (here)[https://developer.apple.com/library/mac/documentation/MacOSX/Conceptual/BPRuntimeConfig/Articles/EnvironmentVars.html]
  • You can set them in your .bashrc/.profile as described for Linux (below)
  • You can set them in each terminal by using export ModelSEEDDatabase=/data/ModelSEEDDatabase

The code in tests/testlp.py uses the example in the documentation (which is also the example in the GLPK documentation)to solve a linear programming problem, and only requires GLPX/GMP and PyGLPK. It does not require any PyFBA code tosolve a simple set of equations. You should check that runs with nosetests tests/testlp.py and it should run a singletest that should pass. If that test does not pass, there is an issue with your installation of GLPK, GMP, or PyGLPK andyou should check that each of them are installed in the correct locations.

There are many more tests in the tests folder and you can run all of them with nosetests tests/. Theyshould all run without an error, and will test different aspects of the PyFBA installation.

If you download and install the code from github, you can also run:

to run all the tests