A Comparison of Procedures for Controlling the False Discovery Rate in the Presence of Small Variance Genes: A Simulation Study

Dan Lin (1) , Ziv Shkedy (1) , Tomasz Burzykowski (1) , Willem Talloen (2) , Luc Bijnens (3)

1 UHasselt - Hasselt University
2 Janssen Pharmaceutical
3 Janssen Pharmaceuticals

The Significance Analysis of Microarrays (Tusher et al. 2001) is widely used in analyzing microarray data. It introduces a fudge factor to adjust test statistic by deflating the large value of test statistics due to the small variances. Lin et al. (2008) pointed out that the fudge factor does not effectively improve the power and control the FDR in the presence of small variance genes. Motivated by those results, we extend our study to compare several methods for choosing the fudge factor and use simulation studies to investigate the power and the control of the FDR of the considered methods.

Mots clés: Physical Sciences

Dan Lin, Ziv Shkedy, Tomasz Burzykowski, Willem Talloen, Luc Bijnens. A Comparison of Procedures for Controlling the False Discovery Rate in the Presence of Small Variance Genes: A Simulation Study. Communications in Statistics - Simulation and Computation, 2009, 38 (10), pp.2111-2122. ⟨10.1080/03610910903249510⟩. ⟨hal-00524862⟩