updated October 28, 2007
Software for Artificial Gene Networks
The generation of artificial gene networks and their use to produce dynamical or steady state data was automated with a series of related programs. The software created is not a polished suite but rather a set of prototype programs in C++ and PERL. Software is not available at this time but may be in the future if it is improved to be a more robust software suite. In particular, a future version to be released would make direct use of our modern simulator COPASI.
The process of creation of a network and derived data works in the following way:
- Step one, carried out by a command line C++ program (actually one program for each specific topology) consists of the creation of a Gepasi file that encodes the network as a set of reactions of trascription and mRNA degradation that follow a specific set of kinetics. (An equivalent file in SBML level 1 format was also created.)
- Step two is to create data using the Gepasi simulator. This is a PERL script that alters the Gepasi model file to specify some experiment (a time course, a gene knockout, etc.), and then run Gepasi to generate the data. (This step uses an undocumented feature of Gepasi which grabs control from other running windows...)
- Step three, in parallel to step two, is to calculate some network statistics, which is done with another PERL script. This uses gnuplot to create a degree distribution plot, and GraphViz to render a drawing of the network.