Figure 5.

A) mRNA expression–protein correlation estimates. The lines with filled and empty circles show all naive pairwise correlation estimates, using mRNAs measured in both data sets only, and the posterior distribution for the correlation, inferred via our structured covariance model, respectively. Dashed vertical lines correspond to mean values. The correlation of the (naive) average protein and mRNA expression levels over measurements is also shown. B) Posterior distribution of the correlation using the technology extension to our model, and discrete technology variance priors on all combinations over the weights 1, 2 and 5. The vertical lines show three weight configurations: the ones with the smallest and largest mean inferred correlation and the equal weighting (all weights equal to 1). C) Posterior distributions of mRNA expression–protein correlations, conditional on exactly one up weighted technology (W = [1, 1, 1, 1, 5]).