GeneChip microarrays-signal intensities, RNA concentrations and probe sequences

GeneChip microarrays consist of hundreds of thousands of oligonucleotide probes. The transformation of their signal intensities into RNA transcript concentrations requires the knowledge of the response function of the measuring device. We analysed the 'apparatus' function of perfect match (PNI) and mismatched (MM) oligonucleotide probes of GeneChip microarrays after changes of the target concentration using the results of a spiked-in experiment. In agreement with previous studies we found that a competitive two-species Langmuir-adsorption model describes the probe intensities well. Each PM and MM probe is characterized by two hybridization constants which specify the propensity of the probe to bind specific and non-specific transcripts. The affinity for non-specific hybridization is on average equal for PM and MM. The purine-pyrimidine asymmetry of base pair interaction strengths, however, causes a characteristic PM-MM intensity difference, the sign of which depends on the middle base of the probe. The affinity for specific hybridization of the PM exceeds that of the MM on average by nearly one order of magnitude because the central mismatched base only weakly contributes to the stability of the probe/target duplexes. For the first time we differentiate between the free energy parameters related to the 64 possible middle-triples of DNA/RNA oligomer duplexes with a central Watson-Crick pairing and a central mismatched pairing. Both the PM and MM probes respond to the concentration of specific transcripts, which can be estimated from the PM and MM probe intensities using the Langmuir-model. The analysis of the PM-MM intensity difference provides at least no loss of accuracy and precision of the estimated concentration compared with the PM-only estimates which in turn outperform the MM-only estimates. The results show that the processing of the PM-MM intensity difference requires the consideration of a background term due to non-specific hybridization, which is, however, reduced by nearly one order of magnitude when compared with the respective background of the PM and MM probes. The calculation of the sequence-specific affinity constants using a positional-dependent nearest-neighbour model opens up the possibility to estimate target concentrations beyond the training set of several hundred of spiked-in probes.