The most statistically significant observations for CD3-stimulated (left) and CD3/CD28-costimulated (right) datasets. Only those features with p ≤ 0.05 for both the correlation coefficient and permutation test are shown. Features are organized by SOM clustering. Sequence-derived parameters are on top, with the others below. Color-coding is as in Fig. 2. The numbers in brackets indicate the genes for which gene ontologies, subcellular localization or structural classes were identified. Amino acids composition parameters were calculated for all the proteins. The actual number of proteins used in any individual analysis varied due to different amounts of missing data in different experiments and at different time points.