• Le laboratoire
• Contacts / Accès
• Membres

• Recherche
• Enseignement
• Séminaire
• Stages

• Publications
• Thèses et HDR
• Logiciels
• Services web

This work results from a collaboration in the Statistics for Systems Biology group between

• the Agroparistech Statistics and Genome group,
• the CNRS/UEVE/INRA Statistics and Genome laboratory,
• the Biometry and Evolutionnary Biology laboratory.

We present a software package based on a new probabilistic model for random graphs called MixNet (Erdös-Renyi Mixture for Networks). This model is based on the hypothesis that real networks are made of classes which show specific connectivity patterns. We compute algorithms to estimate the model's parameters as well as a statistical criterion ICL to select the number of classes.

We analysed various biological networks and summarize the results here.

• The package is written in ANSI C++ and developed on x86 GNU/Linux systems with GCC 4.1.3 It has been successfully tested on Apple Mac OS X.

• Compilation and installation are compliant with the GNU standard procedure.

• The package is free and licensed under the GNU General Public License.
• A man page and a pdf documentation are also available.

The MixNet package provides two programs :

• mixnet implements the variationnal EM algorithm for directed and undirected networks
• mixneto implements various on-line algorithms for large undirected networks

Read their I/O specifications.

A short practise is also available.

• mixnet-1.1.2

The Mixer R package allows the use of the basical options of MixNet software and the post-treatment of the results. It is available at the CRAN.

• nameATbiomserv.univ-lyon1.fr (anti-spam: please replace name by miele)

• Daudin, J.-J., Picard, F., Robin, S., Mixture model for random graphs, Statistics and Computing, 2008.
• Zanghi, H., Ambroise, C., Miele, V., Fast online graph clustering via Erdös-Rényi mixture, Pattern Recognition 41 (3592-3599), 2008.
• Picard, F., Miele, V., Daudin,J-J., Cottret,L., Robin, S., Deciphering the connectivity structure of biological networks using MixNet, BMC Bioinformatics, 2009.