Figure 5

Heatmap depicts the expression of 224 spliced genes with the greatest, at least fourfold mean changes between the two clusters revealed by unsupervised hierarchical clustering. Colour legend indicates the sex, age group, blood collection unit, storage time and library size of each sample. The clusters presented substantial similarity in terms of the expression of spliced junction reads. 62% of spliced platelet RNAs demonstrated less than a twofold expression difference. Only 16% of spliced genes in the dataset (649/3954) demonstrated at least a threefold expression difference, while at least fourfold difference was observed in only 6% of the spliced reads (224/3954). Due to prior clustering, all these observations were statistically significant (FDR corrected q value < 0.05).