Figure 1 .
a) Heatmap showing percentages of overlap in numbers of DEGs between different analyses using different sets of reference genes as well as between tag profiling and microarray analyses. Percentages between different reference sets ranged from 44% to 90% indicating that the choice of the reference set greatly affected what genes were being identified as differentially expressed. Percentages of overlap between microarray analysis and tag profiling were higher with the P. fastigiatum reference sets than with the A. thaliana reference sets in both species (49-60% vs. 21-46%) This indicates that mapping tags to P. fastigiatum is preferable to mapping tags to A. thaliana. All overlaps between microarray and tag profiling analyses were statistically significant (denoted by stars). b) Heatmap showing percentages of contradictory results obtained using different sets of reference genes and between tag profiling and microarrays (= percentage of genes up-regulated in P. enysii with one method and P. fastigiatum with another method). Disagreements between different tag profiling datasets ranged from 0% to 11% indicating the detection of false positives with almost all approaches. All disagreements between tag profiling and microarray analyses were not statistically significant. Also, disagreements did not differ between P. fastigiatum and A. thaliana reference sets (6-12% vs 0-14%). For abbreviations of different sets of reference genes see Aditional file 3 and for calculation of percentages and statistical significance see methods.
Voelckel et al. BMC Genomics 2012 13:322 doi:10.1186/1471-2164-13-322