Using MaxQuant (version 1.2.2.5) (Cox and Mann, 2008), raw data were searched against a database of human protein sequences (ipi.HUMAN.v3.68.fasta), as well as a decoy database of reversed protein sequences, with Arg10 and Lys8 as potential heavy labels. Default parameters were left in place, except the “re-quantify” option was not utilized. More than 300 proteins were identified. The “peptide.txt” output file was used to calculate heavy amino acid incorporation (Geiger et al., 2011). Peptides corresponding to known exogenous contaminants such as keratin and trypsin were first removed. Lysine- and arginine-containing peptides were considered separately. The “ratio H/L” for each peptide was converted to H/(H+L) using the equation H/(H+L) = (“ratio H/L”) / (1+ “ratio H/L”). The median values of H/(H+L) for lysine- and arginine-containing peptides were 0.92 and 0.98, respectively, for each of the four cell lines. These values underestimate incorporation, as a peptide “ratio H/L” cannot be calculated by MaxQuant when there is no detectable signal at the mass-to-charge ratio of the light partner, as would be expected in the case of full incorporation of heavy amino acid.
From Kelly: The only detail I could think to add is to use the R count and K count columns in peptide.txt to focus only on peptides that contain 1 arginine or 1 lysine, to make the calculation straightforward.