Using the GSVD to identify candidate co-expressing genes. This schematic flowchart shows the procedures used to identify a) an overlapping region between the two datasets as well as b) candidate genes probed by the microarray co-expressing with genes of interest from the Q-PCR dataset. Regions A, B and C refer to those defined in Fig. 1. In order to reduce the number of false positives we have repeated the entire procedure a number of times and only examine in detail genes that co-express consistently among these repeats.