ZNova is the core algorithm used by ProMass to deconvolute multiply charged ESI mass spectra. Previous deconvolution algorithms, like the Mann algorithm (1), summed the signals of all masses in a deconvolution range by multiplying the input spectrum by a range of charge values. This type of algorithm works well, but generates a high incidence of artifact peaks that complicate spectral interpretation. ZNova, on the other hand, uses a component deconvolution approach. ZNova determines the charge state of every peak in an ESI mass spectrum by using a scoring method adapted from the ZScore algorithm, as originally described by Zhang and Marshall (2). ZNova evaluates all possible charge states for any particular peak as defined by the input and output mass ranges. A score is assigned based on the intensity of a contiguous series of related charge states (i.e., …z-2, z-1, z+1, z+2…). The candidate charge with the highest cumulative score is assigned the true charge state. Once the charge is determined for a particular m/z peak, ZNova transforms (i.e., multiplies) the m/z peak to obtain its uncharged mass. This process is repeated until all data points in the original spectrum are processed. Unlike many previous algorithms, ZNova has a low incidence of artifact peaks, due to this unique component deconvolution process. ZNova also applies key signal processing techniques that make the approach highly reliable. For example, automatic baseline removal, smoothing, score normalization, and minimum score thresholding are all an integral part of the ZNova deconvolution process.
For more information about deconvolution algorithms
see:
(1) Mann, M.; Meng, C. K.; Fenn, J. B. Anal. Chem.
1989, 61, 1702-1708.
(2) Zhang, Z.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 1998, 9, 225-233.