The flux order relation in experimental data.

Fraction of positive average differences among flux-ordered pairs across the five data types employed, i.e. flux, transcript and protein levels and enzyme kcat values, for the three carbon sources and with a minimum biomass production of 95% of the maximum (i.e., α = 0.95). In the case of data on transcript and protein levels, results are shown for the four different protein cost percentile values, P0—P85 (see main text). Since kcat values can be considered constant under different carbon sources, here we employ the same dataset for the three carbon sources. Experimental p-values from the permutation test are displayed within parentheses. Interactive figures displaying the actual distributions of average data differences are displayed in the online Jupyter Notebook, Section 3.