Deconvolution of Mixture Spectra from Ion-Trap Data-Independent-Acquisition Tandem Mass Spectrometry

摘要

Data-independent tandem mass spectrometry isolates andnfragments all of the molecular species within a given massto-ncharge window, regardless of whether a precursor ionnwas detected within the window. For shotgun proteomicsnon complex protein mixtures, data-independent MS/MSnoffers certain advantages over the traditional data-dependentnMS/MS: identification of low-abundance peptides withninsignificant precursor peaks, more direct relative quantification,nfree of biases caused by competing precursorsnand dynamic exclusion, and faster throughput due tonsimultaneous fragmentation of multiple peptides. However,ndata-independent MS/MS, especially on low-resolutionnion-trap instruments, strains standard peptide identificationnprograms, because of less precise knowledge ofnthe peptide precursor mass and large numbers of spectrancomposed of two or more peptides. Here we describe ancomputer program called DeMux that deconvolves mixturenspectra and improves the peptide identification ratenby ∼25%. We compare the number of identifications madenby data-independent and data-dependent MS/MS at thenpeptide and protein levels: conventional data-dependentnMS/MS makes a greater number of identifications but isnless reproducible from run to run.