Title:Solving the riddle of the bright mismatches: hybridization in oligonucleotide arrays

Abstract: HDONA technology is predicated on two ideas. First, the differential between high-affinity (perfect match, PM) and lower-affinity (mismatch, MM) probes is used to minimize cross-hybridization. Second, several short probes along the transcript are combined, introducing redundancy. Both ideas have shown problems in practice: MMs are often brighter than PMs, and it is hard to combine the pairs because their brightness often spans decades. Previous analysis suggested these problems were sequence-related; publication of the probe sequences has permitted us an in-depth study of this issue. Our results suggest that fluorescently labeling the nucleotides interferes with mRNA binding, causing a catch-22 since, to be detected, the target mRNA must both glow and stick to its probe: without labels it cannot be seen even if bound, while with too many it won't bind. We show that this conflict causes much of the complexity of HDONA raw data, suggesting that an accurate physical understanding of hybridization by incorporating sequence information is necessary to perfect microarray analysis.