Improved protein quantification by using bipartite peptide-protein graphs

In bottom-up proteomics, proteins are enzymatically digested to peptides (smaller amino acid chains) before measurement with mass spectrometry (MS), often using the enzyme trypsin. Because of this, peptides are identified and quantified directly from the MS measurements. Quantification of proteins from this peptide-level data remains a challenge, especially due to the occurrence of shared peptides, which could originate from multiple different protein sequences. The relationship between proteins and their corresponding peptides can be represented by bipartite graphs.

A novel protein quantification method bppgQuant is proposed that is based on the bipartite peptide-protein graphs. The bipartite graphs were constructed from four different quantitative data sets stemming from different organisms (quantitative graphs) as well as theoretical graphs from an in silico digestion of the corresponding protein databases (database graphs). These graphs were characterized and it was shown that a large fraction of protein nodes does not have any unique peptides, which are difficult to quantify and would not be considered by most existing protein quantification methods. This highlights the potential of a novel protein quantification method that can handle those cases.

bppgQuant calculates protein ratios from peptide ratios. It uses the structures of the bipartite peptide-protein graphs to build an equation system and subsequent optimization problem to calculate estimates for the protein ratios. bppgQuant was evaluated on the four above mentioned test data sets containing known protein ratios. Criteria were the MAD and median compared to the expected protein ratios, as well as the number of quantified protein nodes and the ability to distinguish different organisms within one data set or spike-in proteins. In summary, bppgQuant showed good results, especially when unneeded protein nodes were filtered before optimization.