A UNIFIED ALGORITHM FOR DECONVOLUTING ELECTROSPRAY IONIZATION MASS SPECTRAL DATA

摘要

BayesSpray is a novel algorithm for deconvolution of electrospray mass spectral data. We have shown results obtained from mass spectra acquired using a range of instruments under diverse operating conditions. The algorithm provides unified deconvolution of peptide and protein data. The output consists of a list of masses and intensities accompanied by error bars. Other available diagnostics include the evidence (which is used in model comparison), information and mock data for each detected peak. BayesSpray appears to be at least as powerful as MaxEnt 1 and MaxEnt 3, producing significantly cleaner baselines for protein spectra and improved interpretation of the complex data produced by the top down fragmentation of large molecules.