ac9b05272_si_001.pdf (1.01 MB)
A Universal Score for Deconvolution of Intact Protein and Native Electrospray Mass Spectra
journal contribution
posted on 2020-02-25, 21:48 authored by Michael T. MartyThe
growing use of intact protein mass analysis, top-down proteomics,
and native mass spectrometry have created a need for improved data
analysis pipelines for deconvolution of electrospray (ESI) mass spectra
containing multiple charge states and potentially without isotopic
resolution. Although there are multiple deconvolution algorithms,
there is no consensus for how to judge the quality of the deconvolution,
and many scoring schemes are not published. Here, an intuitive universal
score (UniScore) for ESI deconvolution is presented. The UniScore
is the weighted average of deconvolution scores (DScores) for each
peak multiplied by the R2 of the fit to
the data. Each DScore is composed of separate components to score
(1) the uniqueness and fit of the deconvolution to the data, (2) the
consistency of the peak shape across different charge states, (3)
the smoothness of the charge state distribution, and (4) symmetry
and separation of the peak. Example scores are provided for a range
of experimental and simulated data. By providing a means of judging
the quality of the overall deconvolution as well as individual mass
peaks, the UniScore scheme provides a foundation for standardizing
ESI data analysis of larger molecules and enabling the use of ESI
deconvolution in automated data analysis pipelines.