Figure 1.

Upside down pyramid approach of identifying key genes in a condition. As microarray data have lots of false positives (noisy data), it is crucial that layers of gene filtering approaches be applied in order to obtain a set of few genes that may be contributing to the condition or genes that may be signatures of the condition (for clinical diagnostic purposes).

Ray et al. BMC Genomics 2012 13:190   doi:10.1186/1471-2164-13-190
Download authors' original image