PCA interpretation produces 6 files for each principal component (numbered "NN", e.g. "01"), described below:

pcNN-eigenvector.png Plot of the eigenvector (the basis vector) showing relative weighting of conditions (tissues) for a principal component.
pcNN-outliers.txt Tab-delimited text file listing the high and low principal component extreme gene (PCEG) sets for pcNN, including pcNN coefficient, and the gene identifiers and annotations, including: ProbeId, Name, Aliases, LocusLink, Description, Function and Protein Families.
pcNN-outliers.png Scatter plot of probe expression levels projected onto PCn-1 vs PCn space with high PCEG set in red and low PCEG set in blue.
pcNN-condition-groups.txt Tab-delimited text file listing the conditions (tissues) that are up, flat and down for pcNN, ordered by decreasing difference of means.
pcNN-outlier-trajectories-order-original.png Trajectory plots for the high (red) and low (blue) PCEG sets with tissues in the order in which the original data were provided.
pcNN-outlier-trajectories-order-meandiff.png Trajectory plots for high (red) and low (blue) PCEG sets with tissues ordered by decreasing mean differences, and thus grouped by significance (up group at left, flat group in middle and low group at right).

Note: These PCA interpretation files were produced using extremeThresh=0.00001 which produced smaller extreme gene sets.

Note: There is no pc01-outliers.png scatter plot for pc01; see pc02-outliers.png for the pc01 vs. pc02 projection.

Note: The file names refer to the genes as "outliers" for historic reasons; we now refer to these as principal component extreme genes (PCEGs).