Total Mass Difference Statistics Algorithm: A New Approach to Identification of High-Mass Building Blocks in Electrospray Ionization Fourier Transform Ion Cyclotron Mass Spectrometry Data of Natural Organic Matter

• Published in: Analytical chemistry (Washington) Vol. 81; no. 24; pp. 10106 - 10115
• Main Authors: Kunenkov, Erast V, Kononikhin, Alexey S, Perminova, Irina V, Hertkorn, Norbert, Gaspar, Andras, Schmitt-Kopplin, Philippe, Popov, Igor A, Garmash, Andrew V, Nikolaev, Evgeniy N
• Format: Journal Article
• Language: English
• Published: Washington, DC American Chemical Society 15.12.2009
• Subjects: Algorithms; Analytical chemistry; Benzopyrans - chemistry; Fourier transforms; Ions; Mass spectrometry; Models, Statistical; Molecules; Polymethacrylic Acids - chemistry; Polystyrenes - chemistry; Software; Spectrometry, Mass, Electrospray Ionization - methods; Spectrometry, Mass, Electrospray Ionization - statistics & numerical data
• ISSN: 0003-2700; 1520-6882
• DOI: 10.1021/ac901476u

The ultrahigh-resolution Fourier transform ion cyclotron resonance (FTICR) mass spectrum of natural organic matter (NOM) contains several thousand peaks with dozens of molecules matching the same nominal mass. Such a complexity poses a significant challenge for automatic data interpretation, in which the most difficult task is molecular formula assignment, especially in the case of heavy and/or multielement ions. In this study, a new universal algorithm for automatic treatment of FTICR mass spectra of NOM and humic substances based on total mass difference statistics (TMDS) has been developed and implemented. The algorithm enables a blind search for unknown building blocks (instead of a priori known ones) by revealing repetitive patterns present in spectra. In this respect, it differs from all previously developed approaches. This algorithm was implemented in designing FIRAN-software for fully automated analysis of mass data with high peak density. The specific feature of FIRAN is its ability to assign formulas to heavy and/or multielement molecules using “virtual elements” approach. To verify the approach, it was used for processing mass spectra of sodium polystyrene sulfonate (PSS, M w = 2200 Da) and polymethacrylate (PMA, M w = 3290 Da) which produce heavy multielement and multiply-charged ions. Application of TMDS identified unambiguously monomers present in the polymers consistent with their structure: C8H7SO3Na for PSS and C4H6O2 for PMA. It also allowed unambiguous formula assignment to all multiply-charged peaks including the heaviest peak in PMA spectrum at mass 4025.6625 with charge state 6− (mass bias −0.33 ppm). Application of the TMDS-algorithm to processing data on the Suwannee River FA has proven its unique capacities in analysis of spectra with high peak density: it has not only identified the known small building blocks in the structure of FA such as CH2, H2, C2H2O, O but the heavier unit at 154.027 amu. The latter was identified for the first time and assigned a formula C7H6O4 consistent with the structure of dihydroxyl-benzoic acids. The presence of these compounds in the structure of FA has so far been numerically suggested but never proven directly. It was concluded that application of the TMDS-algorithm opens new horizons in unfolding molecular complexity of NOM and other natural products.