Table 1.

Statistical analyses involved in the seven different methods for calculating gene expression.

Methods	Background Correction	Normalization	Core Statistical Analysis	References
AD	None	Invariant Set	Average difference	Affymetrix [2]
MAS5.0	Spatial effect and MM subtracted	Constant	Robust average (Tukey bi-weight)	Affymetrix [2]
MBEI (PM only)	None	Invariant Set	Multiplicative model	Li and Wong [5]
MBEI (PM-MM)	MM intensities are subtracted	Invariant Set	Multiplicative model	Li and Wong [5]
RMA	Global correction	Quantile	Robust linear model (median polish)	Irizarry et al. [4]
PDNN	Model is fitted accounting for background and specific signal	Quantile	Specific and non-specific binding effects are estimated using free energy model	Zhang et al. [7]
GCRMA	Based on probe sequence	Quantile	Robust linear model (median polish)	Wu et al. [8]