Installation

Installation Instructions

System Requirements

A warning for Windows users: fast-tfva has been developed on Linux. If you are using Windows, you should be able to install fast-tfva also. However, it will not work as straight forward as for Linux users and you will probably have to adjust the setup-scripts manually. The code was developed to compile with gcc. It will probably not compile with the microsoft visual studio compiler. If you are not experienced with compiling software on windows using mingw or cygwin, I recommend you to get a machine running Linux to use tfva. If you manage to run fast-tfva on Windows, please tell me how you did it, so I can put the corresponding installation instructions online.

Fast-tfva uses SCIP. Without SCIP, you will not be able to run fast-tfva. For SBML-Support, you need libSBML. A version of libSBML is already packaged with fast-tfva, but you can also use a different version. For Matlab-Support, you can also link in the Matlab libraries. However, Matlab is not open source and hence, the required files cannot be provided.

On university computers it is usually the case that the user does not have root access. Hence, the installation procedure is designed to work also with non-root access. The only exception are the following libraries and programs which are assumed to be installed on your system (should be available on all linux machines by default):

• make
• g++
• bison
• zlib
• readline
• ncurses
• libxml2
• gmp
• libstd
• boost

For ubuntu users

If you need to install some of the above libs, here are the names of the corresponding packages for Ubuntu (probably not comprehensive):

• zlib-bin
• libgmp-dev
• libreadline-dev
• bison
• flex
• libxml2-dev
• libboost-dev
• libncurses5-dev

Installation Instructions

Detailed installation instructions can be found in the README file. For ubuntu/debian users this will be something like (thanks to Elad Noor for writing down those steps):

• sudo apt-get install:
  • zlib-bin
  • libgmp-dev
  • libreadline-dev
  • bison
  • flex
  • libxml2-dev
  • libboost-dev
  • libncurses5-dev
• Download the SCIP tarfile and place in:
  • third_party/res/scipoptsuite-3.0.0.tgz
• Create symbolic links to Matlab:
  • mkdir third_party/matlab
  • ln -s /usr/local/MATLAB/R2012b/extern/include third_party/matlab/include
  • ln -s /usr/local/MATLAB/R2012b/bin/glnxa64 third_party/matlab/bin
• Compile the code:
  • make

Testing Fast-tFVA

For Matlab users the E. coli iAF1260 model is supplied as a default testcase. Simply run make test to check if fast-tfva was installed properly.

For non-Matlab users, there exists no standalone version yet. Hence, no testcase is supplied.