Figure 3. Relative and absolute quantification.

• A
  Ratio of protein mass fractions computed with several protein inference algorithms (x‐axis labels) and mass fractions quantified with spiked‐in labeled peptides (AQUA) for a set of 29 proteins spanning more than 2 orders of magnitude (see Appendix Fig S3, Dataset EV7). The red boxes and symbols represent four proteomics‐based protein inference algorithms (TopPep1/3, iBAQ, and xTop). The blue box and symbols correspond to ribosome profiling‐derived mass fractions. The boxes and whiskers include 50 and 90% of the data, respectively, while the central line represents the median.
• B–E
  Estimated concentration of protein complexes from various protein inference algorithms and ribosome profiling. Individual points are the estimated concentrations for individual proteins in the complex, divided by the number of copies in each complex. Individual concentrations were estimated from the observed protein number fractions assuming a total protein concentration of 3 × 10⁶ proteins/μm³ (Appendix Note S1). The boxes and whiskers include 50 and 90% of the data, respectively, while the central line represents the median.
• F
  The bar chart summarizes the multilinear analysis described in Appendix Note S3. The total height of the bar is the variance in the log‐ratio of proteomics‐based and ribosome profiling‐based mass fractions. Colored components represent the fraction of variance explained by each factor (red, protein size; blue, only one peptide precursor detected; light blue, only 2 peptide precursors detected). TopPep1/3 and xTop protein mass fractions display a bias toward large proteins; TopPep3 and iBAQ systematically underestimate the abundance of proteins with 1 or 2 detected peptides.
• G–J
  Scatter plot of fold change between reference and carbon‐limited conditions (growth rates 0.91/h and 0.35/h, respectively) in mass spectrometry‐based (${\varphi }_i^{\mathrm{C}‐lim}/{\varphi }_i^{\text{ref}}$, y‐axis) and ribosome profiling‐based protein mass fractions (${\rho }_i^{\mathrm{C}‐lim}/{\rho }_i^{\text{ref}}$, x‐axis). The blue line represents equal changes in the two quantities; the dashed lines represent a 2‐fold discrepancy.
• K
  Fraction of proteins showing a > 2‐fold discrepancy in panels (G–J).