PCA: Principal Components

Parameter Information


Sample Selection

The sample selection option indicates whether to perform the analysis on genes or experiments.

Centering Mode

The centering mode selection option determines the type of matrix centering (by mean, median, or none) to be performed before the PCA analysis is run

Algorithm Optimization

This option determines which algorithm to run when clustering by samples. The complete algorithm creates an nxn distance matrix where n is the number of genes. As data sets get large, memory requirements increase exponentially. For most cases, it is a sufficient approximation to calculate the result using the smaller mxm distance matrix where m is the number of samples. This dramatically decreases memory requirements and calculation time.