The filter diagonalization method and its assessment for Fourier transform mass spectrometry

Application of the filter diagonalization method (FDM) to Fourier transform mass spectrometry (FTMS) data is not new. Under certain conditions FDM provides resolution superior to Fourier transform (FT) and was proved to be useful in investigation of space charge phenomena in an ion cyclotron resonance cell (ICR) by O'Connor and Amster research groups. Kozhinov and Tsybin have reported substantial increase in resolution and/or acquisition speed of high-resolution molecular and macromolecular MS data. In light of fundamental difficulty in providing theoretical evaluation of the FDM performance under various spectral and noise conditions, this paper is an empirical investigation aimed at establishing the method's true potentials and areas where it may perform better than currently used technologies. The study was conducted on both synthetic transients and experimental Orbitrap transients. Unlike FT, resolution of FDM depends strongly on noise levels. Consequently, we identify the regimes at which FDM can provide a superior resolution even at moderate signal to noise ratios. Moreover, when individual peaks fail to be resolved either because of the small peak separation or high noise conditions, the FDM solution seems to preserve their cumulative intensity. This preservation of the true intensity seems to be very consistent across rather wide ranges of noise conditions and almost impervious to the peak separation. (C) 2014 Elsevier B.V. All rights reserved.