Figure 3

Correlation plot between the median HS microarray signals in function of the amount of pentapeptides with no match in the human proteome. There were in total 2276, 980, 495, 269, 136, 70, 36, 18, and 4 ‘15-mer’ peptides with, respectively, 0, 1, 2, 3, 4, 5, 6, 7, and 8 5-mer peptide “no matches” in the human proteome. The dotted/dashed lines represent, respectively, the 75th percentile at 2.129, and the median at 1.593 of the complete population. The boxplot median values are, respectively, 1.449, 1.681, 1.820, 2.043, 1,979, 2.661, 2.539, 2.788, and 3.265. All data-points were used in a robust linear regression analysis tool to calculate the slope (0.17), Y intercept (1.55) and the lower (0.14) and upper (0.18) confidence intervals on the slope. The linear regression line is drawn as in interrupted line through the boxplots.