Expert Algorithm for Substance Identification Using Mass Spectrometry: Statistical Foundations in Unimolecular Reaction Rate Theory

Thisstudy aims to resolve one of the longest-standing problemsin mass spectrometry, which is how to accurately identify an organicsubstance from its mass spectrum when a spectrum of the suspectedsubstance has not been analyzed contemporaneously on the same instrument.Part one of this two-part report describes how Rice-Ramsperger-Kassel-Marcus(RRKM) theory predicts that many branching ratios in replicate electron-ionizationmass spectra will provide approximately linear correlations when analysisconditions change within or between instruments. Here, proof-of-conceptgeneral linear modeling is based on the 20 most abundant fragmentsin a database of 128 training spectra of cocaine collected over 6months in an operational crime laboratory. The statistical validityof the approach is confirmed through both analysis of variance (ANOVA)of the regression models and assessment of the distributions of theresiduals of the models. General linear modeling models typicallyexplain more than 90% of the variance in normalized abundances. Whenthe linear models from the training set are applied to 175 additionalknown positive cocaine spectra from more than 20 different laboratories,the linear models enabled ion abundances to be predicted with an accuracyof <2% relative to the base peak, even though the measured abundancesvary by more than 30%. The same models were also applied to 716 knownnegative spectra, including the diastereomers of cocaine: allococaine,pseudococaine, and pseudoallococaine, and the residual errors werelarger for the known negatives than for known positives. The secondpart of the manuscript describes how general linear regression modelingcan serve as the basis for binary classification and reliable identificationof cocaine from its diastereomers and all other known negatives.

Automated summary

This study aims to accurately identify an organic substance from its mass spectrum without analyzing a contemporaneous spectrum of the suspected substance. The first part of the report describes how a theoretical model predicts linear correlations between branching ratios in mass spectra under changing analysis conditions. The model was tested using a dataset of cocaine spectra, achieving accurate predictions even with significant variations in measured abundances. The second part of the report demonstrates how the model can be applied for reliable identification of cocaine and its diastereomers from other known negative spectra.