Table 1.

Examples of normalization options

Normalization option	Description
Median	Measurements are divided by the median of all sample wells in the plate
Shorth	The midpoint of the 'shorth' of the distribution of all sample wells is used for normalization
Mean	Measurements are divided by the mean of all sample wells in the plate
Negatives	Measurements are divided by the median of negative controls in the plate
Percent control	Measurements are divided by the mean of the plate's positive control
Normalized percent control	Measurements are divides by the difference of the plates positive and negative controls
B-score	A two-way (row and column) median polish is applied to each plate
Robust local fit regression	Spatial effects are normalized by fitting a bivariate local regression
Loess regression	Spatial effects are normalized using Loess regression