An Ion Transformation Approach for De Novo Peptide Sequencing via Tandem Mass Spectra

De novo peptide sequencing is especially useful in the identification of unknown proteins. A class of efficient models on de novo peptide interpretation have been proposed recently based on graph theory. However, majority of the graph-based models convert each peak to multiple nodes in the graph due to unknown ion types. This directly results in the decrease of the efficiency and accuracy of peptide sequencing algorithms. To solve these problems, in this paper an ion transformation method is proposed to convert all the other type of ions to b-ions before the construction of spectrum graph in the peptide sequencing. Experiments on both the real QTof MS/MS data and the synthetic MS/MS data demonstrate that the proposed method reduces the size of the graph and improves the efficiency and accuracy of peptide sequencing algorithms.