Interpolational and smoothing cubic spline for mass spectrometry data analysis

• Factors important for mass accuracy of TOF MS were discussed.
• Peak localization bias for peak centroid was shown.
• Equations for histogram spline coefficients were obtained.
• Substantial benefits of histospline over regular spline were not found.
• Some ways of robotic peak detection were proposed.

A special cubic spline for histograms was examined as a possible tool for the processing of mass spectrometry data. Algorithms of interpolating and smoothing cubic splines are described shortly. It was shown that peak apex localizations using splines yield similar results as peak apex determinations using parabolas and centroids for high intensity peaks, while for low intensity peaks, the peak apex localizations using splines are superior. It also was shown that peak localization accuracies for real experimental time-of-flight mass spectral data are higher in case of peak localizations that use apex determinations as compared to peaks determined by centroid or Gaussian approximation of peaks. Some examples of peak detection algorithms using splines are given. A new method of peak intensity calculations is proposed taking into account the baseline. All approaches are directly applicable for the processing of TOF MS data, though specific changes may be desirable for other MS data.

Figure optionsDownload high-quality image (117 K)Download as PowerPoint slide