Despite the success and popularity of microarrays as a high-throughput technology for gene-expression studies, its sensitivity is as yet fairly limited. We have successfully combined the use of PCR-Select cDNA subtraction and Affymetrix GeneChips (AGC) to identify differentially expressed gene markers. Total RNA (totRNA) from combined hippocampus and cerebellum tissues of 2-week-old rat pups maintained for 5 weeks on an n-3 fatty acid (FA) deficient diet supplied to dams was isolated, SMART-amplified, and used for PCR-Select subtraction versus an adequately fed control litter preparation. Subtracted and amplified ds-cDNA end products were fragmented, terminally labeled with biotin-ddUTP and hybridized with RN-U34A AGC. At least 10-fold more potential gene markers with log2(T/D) > or = 1.4 were found versus the traditional AGC technology when the same chip was tested using nonsubtracted targets. Of this set of markers, 30% were robustly validated by real-time relative RT-PCR (rtrRT-PCR) and grouped as "confirmed" markers while the remaining were ascribed as "latent" markers. An improved and universal protocol to provide a rapid assessment for gene profiling in biological specimens is indicated.