Sequenciamento de novo de peptídeos utilizando grafos para espectros de massa multiplex

Abstract

This thesis presents an approach for de novo peptide sequencing (in silico) using multiplex MS2 mass spectra, acquired by a mass spectrometer, containing fragments of more than one peptide in the same MS2 spectrum. A program, named DNbuilder, was developed for de novo peptide sequencing, in which the cost function considers yfragments intensity. The de novo peptide sequencing problem was modeled in graphs using Depth-first search, DFS, as the search algorithm for peptide sequencing over the graph. Within the predefined fragmentation window, monoisotopic charged +2 peptide ions were identified and selected as the MS2 targets. The multiplex methodology consists of the intensity changes of selected peaks from the MS2 multiplex spectrum, including the elimination or reduction of the y or b fragments of the previously identified peptides present in the MS2 multiplex window. The state-of-the-art programs used in the multiplex de novo peptide sequencing tests were Peaks 8, pNovoPlus e Novor 1.3.489, as well as, DNbuilder. A set of spectra from a thyroid sample acquired in 20 m/z window size was used to evaluate the multiplex methodology. The results show the methodology, even if simple, increased the number of correct identified peptide amino acids in multiplex spectra, an evidence that there are ways of de novo sequencing multiplex spectra.