Do young and old protein molecules have the same probability to be degraded? We addressed this question using metabolic pulse-chase labeling and quantitative mass spectrometry to obtain degradation profiles for thousands of proteins. We find that over 10% of proteins are degraded non-exponentially. Specifically, proteins are less stable in the first few hours of their life and stabilize with age. Degradation profiles are conserved and similar in two cell types.Many non-exponentially degraded (NED) proteins are subunits of complexes that are produced in super-stoichiometric amounts relative to their exponentially degraded (ED) counterparts.Within complexes, NED proteins have larger interaction interfaces and assemble earlier than ED subunits. Amplifying genes encoding NED proteins increases their initial degradation.Consistently, decay profiles can predict protein level attenuation in aneuploid cells. Together, our data show that non-exponential degradation is common, conserved and has important consequences for complex formation and regulation of protein abundance.3

Targeted proteomic approaches like selected reaction monitoring (SRM) and parallel reaction monitoring (PRM) are increasingly popular because they enable sensitive and rapid analysis of a preselected set of proteins 1-3 . However, developing targeted assays is tedious and requires the selection, synthesis and mass spectrometric analysis of candidate peptides before the actual samples can be analyzed. The SRMatlas and ProteomeTools projects recently published fragmentation spectra of synthetic peptides covering the entire human proteome 4,5 . These datasets provide very valuable resources. However, extracting the relevant data for selected proteins of interest is not straightforward. For example, developing scheduled acquisition methods (i.e. analyzing specific peptides in defined elution time windows) is complicated and requires adjustments to specific chromatographic conditions employed. Moreover, the number of peptide candidates to be targeted in parallel often exceeds the analytical abilities of the mass spectrometer. In this case, the key question is which peptides can be omitted without losing too much information. Ideally, a method design tool would automatically select the most informative peptides in each retention time window.Until now, none of the available tools automatically generates such optimized scheduled SRM and PRM methods ( Figure S1).