CV Science

Fouzia Moussouni-Marzolf Maitre de Conférences, Univ Rennes1 - Associate Professor in Bioinformatics, Univ Rennes1

Course and current status

10/1999 - present
Associate professor in Bioinformatics at Université de Rennes1 (Maître de
Conférences). Member of INSERM U991 - "Iron and Liver : physiological and pathologoical aspects" Group.        

09/2003 – 02/2004 and 09/2004 – 12/2004
Visiting research position in Berlin, Humboldt University,
funded by the European Science Foundation (ESF)
               
02/1998- 10/1999
Post-doctoral researcher (Research Associate) in Bioinformatics
Information Management Group, Computer Science Department,
Manchester University, UK.

1994-1996
Research & Teaching assistant (ATER) at LIPN.
Université Paris XIII, Paris, France.  

Scientific summary

My early research work focused on the integration of biomedical data from increasingly disparate heterogeneous and rapidly changing life sciences web sources. Interoperability, i.e. linking multiple and distributed information items (gene-function, gene-disease, gene-partner, etc..) was a major issue in bioinformatics, and data integration is still difficult to achieve without human intervention. In this context, we performed the data warehouse GEDAW applied to liver data and QDEX profiler: a Java application able to derive metadata information from data and use them to provide data quality evaluation functions.

Ongoing research work focuses on improving data integration by mining texts of scientific publications. To accelerate data integration on iron relatives, we proceed by text mining of Medline publications related to iron on human models, normal or knockout mice. Indeed, before reaching the biomedical databases, the knowledge that we discover every day on iron homeostasis is present in the literature. In a context of rapidly evolving publications on iron metabolism and with it, increased complexity of iron regulation system at systemic level, and increased complexity of associated pathologies, this work proposes to build a knowledge base on iron homeostasis and ontology associated (qualitative model) by daily mining of scientific publications in Medline.

The integration of recent discoveries lead to better understanding of iron metabolism and better understanding of associated pathologies encountered in patients within the CRFer structure (Centre de Référence des maladies génétiques rares de surcharge en fer). This work is also a collaboration one with Pr. Ulf Leser Humboldt University in Berlin, as part of an Exchange Grant from the ESF-Strasbourg which will begin in April 2011.

Parallel to this, and in partnership with the team LIS at IRISA in Rennes, bioclinical data of CRFer patients will be mined in an exploratory way. The purpose of this search is to extract unexpected associations of patients in the data (hidden patterns) and guide biomedical researcher in the discovery of such patterns. The confrontation of these patterns and the model built from litterature (above) will help formulating new biological hypotheses that can be tested first in experimental models and at mid-term in humans.

All my publications are not indexed by Pubmed. See also the DBLP server (http://www.informatik.uni-trier.de/~ley/db/indices/a-tree/m/Moussouni:Fouzia.html)