AbstractThis thesis contains two principal axes, namely the development of statistical and bioinformatics analyses of data from DNA chip analyses, and their application on biological data relative to the development of metastases induced by hypoxia. The different statistical and bioinformatics analyses are based on work previously performed in our lab for which we brought our contribution, as well as new innovative work performed in the course of this thesis. First, we used a dataset retrieval tool (PathEx) to gather the datasets better suited to our goals. Then, using the results obtained in a benchmarking analysis of single gene statistical analysis methods, we analyzed these datasets using the best suited method. An innovative meta-analysis step performed on the results of the first analyses allowed us to detect differentially expressed genes recurrent in the datasets tested. Then, an over-representation analysis step was performed on the pathways contained in KEGG that allowed to sort the genes obtained at the previous step according to their biological function. In parallel, we developed a new geneset analysis method (FAERI), allowing to statistically test groups of genes biologically related, instead of individual genes. The comparison of the results of the two axes highlights groups of genes supposedly involved in the biological situations tested. The application of these different steps on data related to metastasis and hypoxia allowed us to identify several hundred candidate genes, distributed in tens of pathways. A significant part of these genes are in accordance with recent literature. Moreover, we highlighted the implication of the spliceosome pathway in the metastasis process, which was previously unknown. 20 new candidate genes were identified in this particular pathway. These results suggest the implication of alternative splicing in the development of metastasis. We have developed a statistical and bioinformatics analysis sequence allowing the highlighting of relevant candidate genes, as shown in the analysis of data related to the development of metastasis induced by hypoxia.
|Date of Award||22 Jan 2013|
|Supervisor||Carine MICHIELS (Co-Supervisor), Eric Depiereux (Supervisor), Thierry ARNOULD (President), Jean-Michel DOGNE (Jury), Xavier De Bolle (Jury), Jacques van HELDEN (Jury) & Pierre Sonveaux (Jury)|
Identification de gènes impliqués dans les métastases induites par l'hypoxie, via l'analyse statistique et bioinformatique de banques de damiers à ADN, et confirmation de leur expression différentielle dans des modèles in silico
De Meulder, B. (Author). 22 Jan 2013
Student thesis: Doc types › Doctor of Sciences