RESEARCH

In recent years, a number of sophisticated methods for dealing with ordinary and partial differential equations, inverse and ill-posed problems, learning machines, have been developed in the applied mathematics community.
The availability of these novel methods provides the possibility to address biological problems without precedent and with unmatched complexity, like genome-wide system analysis, structural and dynamic characterization of complex biological networks, high-throughput feature analysis.

Due to the huge amount of data often involved in these applications, we make a significant effort to merge state-of-the-art techniques of numerical analysis with high performance computing, to produce tools able to manage large-scale problems within reasonable computer time.

We envisage that the long term output will be a pool of re-usable modelling components and expertise in their application and interaction that will help elucidate characteristics and relationships of biological processes that would otherwise be impossible to detect and measure.


WHO WE ARE

• Giorgio Fotia
• Fabio Maggio

PROJECTS

The Simulation and Modelling Group has established collaborations with several groups.
Current projects regard include:

• structural and dynamic characterization of yeast gene network (within the RAGNO group)
• high-throughput analysis of bone marrow and blood samples for diagnosis and classification of leukemias

COMPUTATIONAL TOOLS

• network analysis: we developed a set of tool for the statistical eigenvalue analysis of motifs and networks, based on computer tuned linear algebra software and optmized iterative method
• high-troughput feature analysis: we implemented, a supervised classifier based on Proximal Support Vector Machine aimed to large scale data sets, exploiting parallel computing and cache-based linear algebra kernels