Paralog Specificity in cDNA Microarray Experiments

cDNA microarrays have become the technology of choice for profiling the expression of thousands of genes simultaneously in a single sample. However, cross-hybridization between targets and probes that do not correspond must be considered carefully during the design, analysis, and interpretation of these experiments. Here, we examined this potentially confounding issue by competitively hybridizing fragments with similar sequence obtained from two paralogous genes to a custom cDNA microarray that contains probes for only one of these two genes. We found that although a large contiguous portion of these two fragments was 84% identical and only one target was represented on the array, there was very little cross-hybridization of the non-represented target to either the paralogous probes or other (non-similar) features. We conclude that cDNA microarrays are indeed very specific and that promiscuous paralog hybridization adds little noise to the data. More generally, the absence of specific probes for any given gene does not increase the chance for cross-hybridization between non-specific probes and targets, even if sequence similarity is relatively high. PCR products from two paralogs (arginine vasotocin/vasopressin and isotocin/mesotocin/oxytocin) were competitively hybridized onto a glass cDNA microarray platform of the same species. The platform contained probes for only one of the two paralogs in an effort to assess cross-hybridization between paralogous targets and probes.