MaBoSS is a C++ software for simulating continuous/discrete time Markov processes, applied on a Boolean network.
MaBoSS uses a specific language for associating transition rates to each node. Given some initial conditions, MaBoSS applies Monte-Carlo kinetic algorithm (or Gillespie algorithm) to the network to produce time trajectories. Time evolution of probabilities are estimated. In addition, global and semi-global characterizations of the whole system are computed.
- Linux Ubuntu (Ubuntu 4.3.2-1ubuntu11 and higher), RedHat
- MacOS X x86
- Windows with cygwin
Binary distribution
-------------------
To avoid installing compilation tools, we provide binary versions for linux x86, MacOS X x86 and Windows x86:
- linux : MaBoSS-2.0/binaries/linux-x86/MaBoSS
- MacOS X : MaBoSS-2.0/binaries/macos-x86/MaBoSS
- Windows : MaBoSS-2.0/binaries/win-x86/MaBoSS.exe
Important notes on the Windows version:
- to execute MaBoSS.exe, cygwin must be installed (http://www.cygwin.com/)
- because of the cygwin emulation, the windows version is very slow (about 4 times slower than the linux and Mac OS versions). We urge you to run MaBoSS on linux or Mac OS X if possible.
All these binary versions are provided "as is", they may not work on your OS. In such a case, you need to compile MaBoSS.
Compilation
-----------
Tools needed
- gcc: version 4.0.1 or higher
- bison: version 2.3 or higher
- flex: version 2.5.35 or higher
- cygwin needed on Windows
cd MaBoSS-2.0/src
make install
The executable file will be located in MaBoSS-2.0/pub and is named MaBoSS.
This compiled version supports up to 64 nodes per network.
If you need more nodes per network, you have to add an extra hint to the compilation command, for instance to compile a version supporting up to 100 nodes:
make MAXNODES=100 install
The executable file will also be located in MaBoSS-2.0/pub and will be named MaBoSS_100n.
Notes:
- if you manage only networks with up to 64 nodes, we recommend you to use the default compiled version as for networks with more than 64 nodes, the implementation is very different and will be slower.
- generally speaking, a version compiled with a given number of nodes will be slower and will use more memory than a version compiled with a lesser number of nodes.
Usage
-----
cd MaBoSS-2.0/pub
./MaBoSS --version
MaBoSS version 2.0 [networks up to 64 nodes]
./MaBoSS_100n --version
MaBoSS version 2.0 [networks up to 100 nodes]
The usage is described in the reference card MaBoSS-2.0/doc/MaBoSS-RefCard.pdf
Examples
--------
The directory MaBoSS-2.0/examples contains two examples:
- ToyModel
- p53_Mdm2
To test an example (for instance p53_Mdm2):
- change to the directory containing the model (MaBoSS-2.0/examples/p53_Mdm2)