Table 1.

Graphical outputs and their functions for the RACE Quality Checks & Normalization tool

Output	Function
CEL intensity images	Detect spatial intensity artifacts
Boxplots of raw PM probe intensities	Display PM probe intensity distribution for selected array; compare overall brightness of selected array
Density distribution of the raw PM probe intensities	Check the intensity density for selected arrays; compare the densities of selected arrays
Boxplots of the normalized PM intensities	Assess the success of the normalization
5′ to 3′ feature intensity plot (‘RNA digestion plot’)	Detect a bias in probe intensities; identify outlier arrays with deviating biases
PLM pseudo images	Assess spatial distribution of weights derived during robust linear model probe-level fit; detect obscure/dark array regions with low weights; assess the spatial distribution of residuals; detect obscure/dark array regions with high positive or negative residuals
NUSE boxplots	Identify arrays where the standard errors for gene expression estimates from PLM fit are overall larger relative to other arrays
RLE boxplots	Identify arrays where relative log expression compared with a median array are larger than other arrays
Pair-wise scatter plots	Assess similarities and differences in expression values measured on two arrays; identify outliers arrays
Correlation matrix plot	Detect homogeneous groups of arrays; identify outlier arrays
Sample cluster	Find subgroups of similar samples