Fig. 1.

Dataset 3: Spike-in benchmark—ubiquitination—label-free.A, four mixtures (i.e., conditions) were created with varying amounts of human lysate, background E. coli lysate, and human spike-in Ub-peptide mixture. Unmodified peptides from human lysate were viewed as the global proteome. Background E. coli lysate were used to equalize total protein levels. Fifty heavy-labeled KGG motif peptides from 20 human proteins were spiked into the mixed background of the lysates. Quantitative changes in protein and site abundance of these 20 human proteins were the target of the benchmark. B, we distinguished the unadjusted changes (i.e., changes in the abundances of the modified peptides) and the protein-level adjusted changes of (i.e., changes in the abundances of the modified peptides relative to the changes in the abundances of the human lysate). “Unadj. true log₂FC” are the log-ratios of the abundances of the spiked peptides between each condition. “Adj. true log₂FC” was calculated by determining the ratios of the abundances of the spiked peptides and human lysate between each condition and then adjusting the ratio of the spiked peptides by the human lysate, similarly to (Equation 2).