XhybCasneuf package

This package contains the data that was generated by and investigated in our study, ’The Effect of Cross-Hybridisation on Expression Correlations Inferred from Microarrays’ (manuscript submitted). In the first part, we investigated the relationship between reporter-to-transcript sequence similarity and correlation of expression signals. We assessed the extent to which off-target reporters in probe sets, i.e. reporters that (partially) align to another transcript than the one intended, influence the expression correlation of the intended and off-target probe set. For a given probe set X, intented to target gene X and a potential off-target gene Y, the variables were calculated as followed: First X’s reporters’ sequences were aligned to the transcript of Y. To quantify the potential off-target affinity of probe set X to Y, the 75th percentile Q75 XY was then calculated of these alignment scores {a1, . . . , an}. We obtained the expression correlation measure by calculating the Pearson correlation coefficient between the intensity values of X and Y in 14 different plant tissues. These data are a subset of the AtgenExpress project data [1, 2]. We illustrated our findings with three detailed examples of cross-hybridisation. We also ran a simulation to demontstrate the effect of individual reporters on summarisation. Our study shows that numerous probe sets on a widely used commercial array platform contain off-target reporters, and many probe sets show a signal pattern that is highly similar to that of unintended transcripts. In addition, a positive correlation is revealed between off-target alignment score of different reporters and the magnitude of their off-target correlation. We evaluated the conventional probe set design, as defined by the array manufacturer (Affymetrix CDF), and compared it to a custom-made CDF. We demonstrate that careful reporter mapping alleviates cross-hybridisation effects to a substantial extent. This analysis was conducted on the ATH1 Affymetrix GeneChip for Arabidopsis thaliana. This package contains, for both the Affymetrix and the custom-made CDF, the data of probe set pairs with a off-target sensitivity score of ≥ 55. Load the library: