Figure 1

Identification of unannotated extended probe sets. A. At the core of this analysis is our murine mRNA expression database comprising 22 different tissues. Unannotated probe sets binding putative 3' UTR extensions were identified as shown in the flow chart. The number of RefSeq transcripts retained after each filtering step is shown, finally resulting in 922 transcripts corresponding to 845 unique gene symbols for which 3' UTR extensions are predicted. B. Histogram with Pearson's correlations. The red histogram depicts correlations between the 1849 probe set pairs selected as shown in panel A. The blue histogram depicts correlations between random pairs of probe sets from the same microarray platform. As a cut-off for statistically significant co-expression, a Pearson correlation of 0.6 was chosen, resulting in an estimated false positive rate of 2%.