Olivier Gascuel

Méthodes et Algorithmes pour la Bioinformatique
LIRMM-CNRS, 161 rue Ada,
34392 - Montpellier - FRANCE
Tel. 33 (0) 4 67 41 85 47
Fax. 33 (0) 4 67 41 85 00
gascuel@lirmm.fr

· Recherches - research

· Laboratoire - lab

· Famille - family

Research interests

My main research interest is in the field of evolution and phylogeny. My focus is on the mathematical and computational tools and concepts, which form now an essential basis of evolutionary studies (2005 book, 2007 book).

I am working on algorithms for phylogenetic inference. With Alexis Criscuolo we conceived the SDM distance-based method to combine gene trees into a species supertree (Syst Biol 2006), and the PhyD* algorithms to build trees from distance matrices with missing entries (BMC Bioinformatics 2008). With Stéphane Guindon we designed the PhyML software for maximum-likelihood tree inference (Syst Biol 2003, Highly Cited!). This software has been refined and enriched with several tools, most notably the aLRT branch supports (Syst Biol 2006). PhyML 3.0 paper just appeared (Syst Biol 2010), describing these novelties and extensive benchmarking with a number of DNA and protein data sets.

I am working on probabilistic models of sequence evolution, most notably proteins. With Quang Si Le we recently proposed several evolutionary models based on mixtures (Phil Trans Roy Soc B 2008) or site partitions using available structural information (Syst Biol 2010). We also designed a new estimation method for amino acid replacement matrices and proposed the LG matrix (Mol Biol Evol 2008). With Mike Steel we explored the impact of substitution models on the predictability of ancestral sequences (Math Biosciences 2010).

During the last years, I turned part of my activities toward pathogens, most notably Plasmodium faciparum which is the main agent of malaria, responsible for more than one million deaths per year. Our aim was to decipher the function of its genes, most of them (60%) being fully unknown, while they could be potential targets for new drugs or vaccines. We used postgenomic data (mostly transcriptomes) and a statistical learning approach to propose functional predictions for a number of its genes (BMC Bioinfoirmatics 2008); these predictions are compiled in the PlasmoDraft database. We also refined the annotation of its proteins in structural domains (Bioinformatics 2009).