Comparison of algorithms and databases for matching unknown mass spectra

Abstract

The most used algorithms for the identification of electron-ionization mass spectra are INCOS and probability based matching (PBM). For unknown spectra of high purity, approximately 75% of rank 1 answers are correct for both algorithms, matched against the National Institute of Standards and Technology 62,235 spectrum database. With matching criteria that retrieve 50% of the possible correct answers from the Wiley 228,998 spectrum database, 54% of the PBM and 42% of the INCOS answers are correct; for 85% purity unknowns, 48% and 27% are correct. For an unknown spectrum of two compounds, neither was reported in the first three INCOS answers; eight of the first ten PBM answers identify both components.

Quantitative Comparison of Tandem Mass Spectra Obtained on Various Instruments

“…Now emphasis is shifting to automated procedures: (1) comparison with and search of data banks, and (2) comparing experimentally obtained and theoretically expected mass spectra. Data bank search has long been in use, especially for electron impact spectra [1][2][3]. Comparison to expected (theoretical) tandem mass spectra is the basis of proteomics [4], even if comparisons use predominantly fragment ion masses only, with no or limited use of ion abundance.…”

Investigation of pyrolysis behavior of carbofuran by pyrolysis-gas chromatography–mass spectrometry

“…The PBM algorithm [21,22], which used a reverse search to verify that peaks in the reference spectrum were present in the unknown spectrum, was used to match the detected mass spectra. The spectral similarity was measured by reverse match factor (R.Match).…”

“…In Py-GC-MS, the pyrolysis components are separated and identified by matching the measured mass spectrum against the standard mass spectra library. The commonly used method is the probability-based match (PBM) algorithm [21,22]. For the complex mixtures, the PBM results can be determined with the aid of correlation of boiling point (bp) and retention behavior [23,24].…”

Neutral losses: A type of important variables in prediction of branching degree for acyclic alkenes from mass spectra

“…Generally, the widely available identification method is mass spectral library searching [14] using various search algorithms including dot-product function [15] and probability based matching (PBM) [16][17][18]. In fact, these libraries serve not only mass spectral searching but also data mining and discovery of mass spectral characteristics for structural elucidation [19].…”