Premium
Towards a full reference library of MS n spectra. Testing of a library containing 3126 MS 2 spectra of 1743 compounds
Author(s) -
Milman Boris L.
Publication year - 2005
Publication title -
rapid communications in mass spectrometry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.528
H-Index - 136
eISSN - 1097-0231
pISSN - 0951-4198
DOI - 10.1002/rcm.2131
Subject(s) - chemistry , spectral line , mass spectrum , analytical chemistry (journal) , mass spectrometry , chromatography , physics , astronomy
A library consisting of 3766 MS n spectra of 1743 compounds, including 3126 MS 2 spectra acquired mainly using ion trap (IT) and triple‐quadrupole (QqQ) instruments, was composed of numerous collections/sources. Ionization techniques were mainly electrospray ionization and also atmospheric pressure chemical ionization and chemical ionization. The library was tested for the performance in identification of unknowns, and in this context this work is believed to be the largest of all known tests of product‐ion mass spectral libraries. The MS 2 spectra of the same compounds from different collections were in turn divided into spectra of ‘unknown’ and reference compounds. For each particular compound, library searches were performed resulting in selection by taking into account the best matches for each spectral collection/source. Within each collection/source, replicate MS 2 spectra differed in the collision energy used. Overall, there were up to 950 search results giving the best match factors and their ranks in corresponding hit lists. In general, the correct answers were obtained as the 1 st rank in up to 60% of the search results when retrieved with (on average) 2.2 ‘unknown’ and 6.2 reference replicates per compound. With two or more replicates of both ‘unknown’ and reference spectra (the average numbers of replicates were 4.0 and 7.8, respectively), the fraction of correct answers in the 1 st rank increased to 77%. This value is close to the performance of established electron ionization mass spectra libraries (up to 79%) found by other workers. The hypothesis that MS 2 spectra better match reference spectra acquired using the same type of tandem mass spectrometer (IT or QqQ) was neither strongly proved nor rejected here. The present work shows that MS 2 spectral libraries containing sufficiently numerous different entries for each compound are sufficiently efficient for identification of unknowns and suitable for use with different tandem mass spectrometers. Copyright © 2005 John Wiley & Sons, Ltd.