Séminaire de biostatistique, Hôtel-Dieu de Québec, 2 septembre 2011

le 29 août 2011 à 10:01
Jean-Francois Plante

Séminaire de biostatistique organisé par le Centre de recherche du CHUQ à l’Hôtel-Dieu de Québec

Vendredi le 2 septembre 2011, 11:30
Amphithéâtre du CRCEO, L’Hôtel-Dieu de Québec, 6 rue McMahon

Statistical Genetics in Cancer Research: Methodology and Application
Dr. Wei Xu, Assistant Professor, Division Biostatistics, Princess Margaret Hospital, Toronto

Research in human genetics has recently focused on investigating more complex diseases such as most of the cancer diseases that may be controlled by both genetic and environmental factors, possibly interacting together. However, identify the disease gene or multiple genes and associated environmental factors on complex human diseases can be difficult since the relationship between a particular gene or genes and the disease trait(s) is often unclear. Currently many clinical cancer studies work on finding significant prognostic factors including clinical and genetic factors. But few have been done in exploring combined effect of those factors including their interactions. Since the combined and interactive effects can be complicated and in a non-linear way on the clinical outcome such as OS, DFS, and toxicity. Non-linear statistical algorithms such as recursively partition (RP) model can be applied and to identify homogeneous subgroups with high risk or suitable treatment. The RP model has the potential to identify previously unknown gene-environment interactions, and the clinically practical utility and potential this new methodology holds have been shown in a randomization clinical trial on head and neck Cancer.