RESEARCH LINES:

• METABOLIC SYSTEMS
• ANGIOGENESIS
• SIGNAL TRANSDUCTION
  AND CANCER
• ProFunc: (PRO)TEIN
  (FUNC)TION GROUP

EDUCATION

JOB OPPORTUNITIES

PHOTO

Prediction, modelling and protein networks topological analysis of the entire H. sapiens proteome

The group has developped two kind of interaction networks amongst proteins for the whole H. sapiens proteome. We term them KnowledgeGram (KG) and PredictoGram (PG) by the source of the built-on data. KG networks comprise the information about proteins pairwise functional association yielded from databases of Biological "Knowledge" and/or experimental, in the meantime PG networks form ab-initio predictions from protein interactions obtained by several bioinformatics methods deployed and developped within the group. We have demonstrated that PG networks, based on predictions, resemble the topological features of biological experimental KG and contain key functional information the most not gathered yet (80%-90%) in experimental evidences-based networks. Those networks were used in predicting novel functions of the histamine H4R receptor in brain.

Modelling and evolutive analysis of proteins complex for the Escherichia coli and Saccharomyces cerevisiae proteomes

We work on the high precision modelling of proteins complex for the Escherichia coli and Saccharomyces cerevisiae proteomes based on proteins complex currently accepted by the scientific community and complexes inferred from interactions proteins High-Throughput assays. We observed remarkable differences in the way that proteins complex in both proteomes evolve, which suggests different evolutive approaches between prokaryotes and eukaryotes.

Prediction of novel proteins involved in the H. sapiens mitotic spindle construction

By the integrated platform for the mitotic spindle proteins prediction, developped in collaboration with several European groups within the network of excellence in systems Biology ENFIN (Experimental Network for Functional INtegration), we carried out the prediction of mitotic spindle proteins experimentally validated above 80% of being true against the much lower 35% obtained in former approaches.

FuncNet: Bioinformatics Methods of Integration for the High-Throughtput prediction of protein functions

In collaboration with the ENFIN network of excellence as well, we deployed FuncNet. FuncNet is a informatic web service for the proteins function prediction integrating several well-established predictive methods. Our group deployed the algorithm and mathematical integration procedure for this web service. FuncNet enables the users from a front-end to integrate different results obtained in turn from several predictive algorithms in an overall query process, increasing at the meantime its predictive power by the combination and/or integration of the different methods. For further information see the www.funcnet.eu link.

3D modelling and computational simulations

We apply the structural proteins modelling techniques and simulations of the molecular dynamic to different systems of our interest with the aim of characterizing the relationship structure-function in them. As recent examples of analysed systems we enlarge the interaction complex of the 14-3-3 protein with KSR1, the mammalian histidine descarboxylase, the histamine H4R receptor and the plant progesterone 5²-reductase.