Affymetrix gene chips use small oligo nucleotide probes to measure gene expression. This technology has disadvantages when applied to species other than those used to develop the gene chips. For one, DNA sequence divergence leads to mismatches between a gene's RNA and the probe for that gene on the chip. This can lead to inaccurate measurements of gene expression.
Affymetrix arrays contain 14 pairs of 25-nucleotide probes for each gene. One of this pair is a 'perfect' match (PM) to the D. melanogaster reference sequence. The other differs only at the 13th base and is the mismatch probe (MM). In D. simulans and D. yakuba, some MM probes exactly match the genomic sequence of D. simulans and D. yakuba because of divergence between these species and D. melanogaster. These paired sets of probes where the PM is a one base mismatch and the MM is an exact match are ideal for determining the effects of a single base mismatch on hybridization intensity. We identified these informative probe pairs. We then compared the normalized hybridization intensity for each of these 'exact match' MM probes to the data from the now imperfect PM probes.
Our data shows that single-base mismatches typically lower hybridization intensity by a third and by as much as 10 to 20 fold. On average for both D. simulans and D. yakuba the effect of a single base mismatch was about a 30-36% decline in hybridization intensity. The range of this effect, however, is considerable (see figure 1). It is worth noting that 19% of single base pair mismatches led to a two fold difference in hybridization intensity. One would expect that two-base mismatches would have an even greater effect.