Premium
Statistical approach to establish equivalence of unabbreviated mass spectra
Author(s) -
Bodnar Willard Melissa A.,
Waddell Smith Ruth,
McGuffin Victoria L.
Publication year - 2013
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.6759
Subject(s) - chemistry , confidence interval , replicate , mass spectrum , statistics , standard deviation , analytical chemistry (journal) , equivalence (formal languages) , spectral line , homologous series , fragmentation (computing) , chromatography , mass spectrometry , mathematics , physics , organic chemistry , discrete mathematics , astronomy , computer science , operating system
RATIONALE In many legal and regulatory applications, mass spectral comparison of an unknown or questioned sample to a reference standard or database is used for identification; however, no statistical confidence level or error rate is determined. Therefore, a simple and rapid method to establish the statistical equivalence of mass spectra is needed. METHODS The standard deviation of the abundance at each m/z ratio was determined from replicate measurements or from a statistical model. These standard deviations were used in an unequal variance t ‐test to compare two spectra at every m/z ratio over the entire scan range. If determined to be statistically indistinguishable at every m/z ratio, the random‐match probability (RMP) that the specific mass spectral fragmentation pattern occurred by chance was calculated. RESULTS n ‐Alkane and alkylbenzene standards of varying concentrations were analyzed on the same instrument at different ionization voltages. Using the proposed method, replicate spectra were successfully associated at the 99.9% confidence level, with RMP values less than 10 –29 . Despite the similarity in fragmentation patterns, spectra were distinguished from others in the homologous series. Moreover, the n ‐alkane spectra were appropriately associated to and discriminated from those in a standard reference database at the 99.9% confidence level. CONCLUSIONS A simple and rapid method to assign statistical significance to the comparison of mass spectra was developed and validated. This method may be useful for legal and regulatory applications, such as the identification of controlled substances, environmental pollutants, and food and drug contaminants. Copyright © 2013 John Wiley & Sons, Ltd.