Laurent Bréhélin

ma bobine

Méthodes et Algorithmes pour la Bioinformatique (MAB)
Laboratoire d'Informatique Robotique et Microélectronique de Montpellier (LIRMM)
161 rue Ada
34095 Montpellier Cedex 5 - France

b r e h e l i n @ l i r m m . f r
tel: +33 467 41 86 71
fax: +33 467 41 85 02




I am a CNRS researcher at the LIRMM. Since 2015 I am also the head of the MAB team.



I am interested is the development of statistical and machine learning methods for functional annotation of genomes, with a particular focus on human and Plasmodium falciparum (malaria) genomes.

Part of my work is dedicated to the annotation of protein domains, which constitute the functional units of the proteins. For this, we proposed several methods to improve the sensibility of domain detection with Hidden Markov Models (Bioinformatics 2009, BMC Bioinformatics 2012, PLOS One 2014) and for identifying new domain families not yet referenced in protein domain databases PLOS CB 2018.

During the last years, I turned most of my activities to the study of the genomic signals responsible for gene expression regulation. With colleagues from IGMM and IMAG, we develop machine learning approaches for identifying the cis regulatory elements associated with gene expression, transcription factor binding sites and other chromatin marks (Genome Biology 2012, PLOS CB 2018b , BMC Genomics 2019, BioRXiv 2020a, BioRXiv 2020b.

Publications

See my list of publications in HAL.

Databases and software

  • DExTER is a program written in python to identify Long Regulatory Elements, i.e. genomic regions of several dozen bps whose nucleotide content is linked to gene expression (BioRXiv 2020a)
  • TFcoop is a method in R to infer the TF combinations involved in the binding of a target TF in a particular cell type (BMC Genomics 2019).
  • blast2domains is a set of scripts written in Python to identify new families of protein domains on the basis of a blast analysis (PLOS CB 2018).
  • EuPathDomains is an extended database of protein domains in several eukaryotic pathogens (Ghouila et al. 2010). The EuPathDomains database gathers known Interpro domains occurrences and new Pfam domain occurrences found by the CODD procedure, which improves domain detection sensitivity by exploiting co-occurrence properties (Terrapon et al. 2009).
  • PlasmoDraft is a database of Gene Ontology (GO) annotation predictions for the genes of Plasmodium falciparum (Bréhélin et al. 2009). Predictions of PlasmoDraft have been produced by a Guilt By Association Approach applied on transcriptome, proteome, and interactome data, and are thus complementary to predictions achieved by sequence homology.
  • Stability.so is an R package that evaluates the stability of clusters derived from a hierarchical clustering of gene expression data when repeated measurements are available (Bréhélin et al. 2008).

PhD students and postdoctoral fellows

  • Yulia Rodina (postdoc)
  • Raphaël Romero (PhD)
  • Christophe Menichelli (PhD 2019)
  • Jimmy Vandel (postdoc 2018)
  • May Taha (PhD 2018)
  • David Martin (postdoc 2013)
  • Mathieu Lajoie (postdoc 2012)
  • Amel Ghouila (PhD 2014)
  • Nicolas Terrapon (PhD 2010)