Raffronto di Vari Metodi per l’Identificazione dell’Espressione Genica Differenziale, un’Applicazione con Oligonucleotidi alla Soppressione del Gene Antiapoptosi Survivina

A very common problem that is adressed using microarray technology is that of identificating genes that are differentially expressed between two or more classes of samples. For this purpose many methods have been proposed and made available to a wide range of users by supplying related software. Refinements of familiar tools such as the t-test are used to take the multiplicity problem into account (Dudoit et al. 2002), and to obtain a reliable estimate of the variance of gene expression (Tusher et al. 2001), as well as model based methods, both in a frequentist (Kerr and Churchill 2001) and Bayesian framework (Gottardo et al. 2002). These methods imply different hypotheses and their use on data can lead to substantially different results. The purpose of this work is to evaluate and compare, through a case study, results from different methods for identifying differentially expressed genes.