Affichage des résultats 81 à 100 sur 1566 au total
Thèse de Béatrice Donati le mercredi 12 novembre 2014 à 10 h - amphithéâtre Physique Nucléaire bâtiment Paul Dirac (Doua)
(travail en collaboration avec A. Arribas-Gil, Univ. Carlos III, Madrid) Dans cet exposé, je parlerai tout d'abord de l'alignement statistique de 2 séquences, qui permet d'aligner ces séquences sans choisir a priori les paramètres de l'alignement. Ces paramètres sont en fait déterminés par maximum de vraisemblance, dans un modèle d'évolution de ces deux séquences. Dans un second temps, je montrerai commen on peut modifier l'algorithme d'alignement statistique pour prendre en compte la non indépendance dans l'évolution des sites de la séquence. J'illustre ces résultats sur une paire de séquences "jouet" (alpha-globine humaine HBA1 et pseudo-gène HBPA1).
-
-
-
-
Thèse de Laurent Modolo le lundi 1er décembre 2014 à 14 h - Salle de conférence Bibliothèque Universitaire (Doua)
L'exploration numérique de modèles, basée sur la simulation, s'est beaucoup développée depuis une quinzaine d'années. L'analyse de sensibilité globale, en particulier, permet d'étudier et comparer l'influence de paramètres ou variables d'entrée incertains sur les sorties d'un modèle, en considérant des plages de variation de ces facteurs plutôt que des dérivées autour de valeurs ponctuelles. Les principes de l'analyse de sensibilité globale seront présentés, ainsi que quelques méthodes telles que Sobol' et FAST. L'approche sera illustrée sur un modèle théorique développé pour étudier la propagation d'un champignon sur un paysage agricole.
Thèse de Ghislain Durif le mardi 13 décembre 2016 à 13 h 30 - salle de conférence BU (La Doua)
-
-
Thèse de Hélène Lopez le mercredi 15 février 2017 à 14 h, salle Fontannes bâtiment Darwin D (La Doua)
Thèse de Thibaut Jombart - 14 novembre 2008 - 14h00 - Amphitéâtre du CNRS
Structural variation in general, and copy number variants (CNV) in particular, has emerged as an important source of genetic variation. The genetic history and the extraordinary morphological, physiological and behavioral variation of dogs, make them an ideal mammal in which to study the effects of CNV on biology and disease. The dog genome revealed the existence of more than one thousand of CNV that overlap ≈ 400 genes, which are enriched for defense/immunity, oxidoreductase, protease, receptor, signaling molecule and transporter genes. Furthermore, CNV can have significant impacts on a wide range of phenotypes including breed-definiting traits and showed to be appropriate markers to analyze genetic relationships between dog populations. This finding implies that most of the surveyed CNVs were present in the pool of canine breed founders. In order to understand the ancestral dog genome organization, we designed a high density custom 720K probes NimbleGen aCGH chip based on all known dog CNV and segmental duplication and genotyped 15 wolves from 11 populations, with a wide distribution (including Europe, Asia and America), 5 dogs (Dingo, Basenji, Beagle, Boxer and Dachshund) and three outgroups (red wolf, coyote and golden jackal). The dataset analyzed in this study allow us to identify selected CNV during early dog domestication.
When a patient contract cancer, the diagnosis and the treatment options are decided based on phenotypic factor such as the primary organ and the pathology of the tumors. However, recent discoveries have shown than the molecular pathogenesis of the tumor may be a better indicator of the prognosis and the adequate therapeutical targets, underlining the need for a better understanding of the genomic characteristics of the disease. In order to solve this issue, initiatives such as TCGA have sequenced large groups of disease samples and used the size of the cohort to distinguish between driver and passenger events. We propose to adopt the same approach to analyze liquid tumors, where widespread heterogeneity within the tumor as well as between patients adds even more noise to the measurements. We will show the software infrastructure we built in order to track and analyze the samples as well as two examples of such studies. The first cohort we will discuss is a cohort of 53 patients with acute myeloid leukemia (AML). AML is the most common type of leukemia in adult, and the is still poorly understood. All these patients achieved complete remissions after standard chemotherapy and later suffered from relapse. We performed whole exome sequencing of germline tissue, primary sample as well as relapsed sample for each of these patients and will show analysis of this data compared to the knowledge that we have of the disease. The second cohort we will discuss is a cohort of 189 adult patients with B-cell acute lymphoblastic leukemia (B-ALL). B-ALL is the most common leukemic malignancy in the pediatric population but most of the knowledge that we have from the adult form of the disease is actually deduced from the childhood form. We will present here a integrated and targeted DNA/RNA sequencing solution that allowed us to extract short events (point substitutions and short insertions and deletions) as well as copy number aberrations, rearrangements and fusion events. We will use these two cases to discuss common problems that arise when analyzing genomic data as well as potential solutions."
This communication will have two distinct aims. Firstly, it is intent to introduce my work in the area of philosophy of science to the members of the Laboratory I joined last November. I will briefly present my two main research themes, namely 1° the difficulty to define a scientific object such as animal behavior, and 2° scientific reasoning in biology, as they have been introduced in the PhD thesis I defended last year (Epistémologie historique de l'étude du comportement animal). Secondly, this communication also aims to convince the audience about the relevance and the potential utility of philosophy of science for the practice of science.