Fabienne Thomarat

Learn More
Microsporidia are obligate intracellular parasites infesting many animal groups. Lacking mitochondria and peroxysomes, these unicellular eukaryotes were first considered a deeply branching protist lineage that diverged before the endosymbiotic event that led to mitochondria. The discovery of a gene for a mitochondrial-type chaperone combined with molecular(More)
The DNA sequences of the 11 linear chromosomes of the approximately 2.9 Mbp genome of Encephalitozoon cuniculi, an obligate intracellular parasite of mammals, include approximately 2000 putative protein-coding genes. The compactness of this genome is associated with the length reduction of various genes. Essential functions are dependent on a minimal set of(More)
Microsporidia are unicellular eukaryotes living as obligate intracellular parasites. Lacking mitochondria, they were initially considered as having diverged before the endosymbiosis at the origin of mitochondria. That microsporidia were primitively amitochondriate was first questioned by the discovery of microsporidial sequences homologous to genes encoding(More)
We sequenced the complete control region (CR) and adjacent tRNAs, partial 12S rRNA, and cytochrome b (over 3100 bp) from eight individuals of Madeiran wall lizards, Lacerta dugesii, from four distinct island populations. The tRNAs exhibit a high degree of intraspecific polymorphisms compared to other vertebrates. All CR sequences include a minisatellite(More)
To tackle segmentation problems on biological sequences, we advocate the use of a hybrid architecture combining discriminant and generative models in the framework of a hierarchical approach. Multi-class support vector machines and neural networks provide a set of initial predictions. These predictions are postprocessed by classifiers estimating the class(More)
Support vector machines, let them be bi-class or multi-class, have proved efficient for protein secondary structure prediction. They can be used either as sequence-to-structure classifier, structure-to-structure classifier, or both. Compared to the classifier most commonly found in the main prediction methods, the multi-layer perceptron, they exhibit one(More)
Most of the state-of-the-art methods for protein seconday structure prediction are complex combinations of discriminant models. They apply a local approach of the prediction which is known to induce a limit on the expected prediction accuracy. A priori, the use of generative models should make it possible to overcome this limitation. However, among the(More)
  • 1