Signal Processing for LC-MS & MS/MS

Denoising of LC-MS data by MEND

A new denoising and peak picking algorithm (MEND - matched filtration with experimental noise determination) for analysis of LC-MS data has been developed. The algorithm minimizes both random and chemical noise in order to determine MS peaks corresponding to sample components. Noise characteristics in the data set are experimentally determined and used for efficient denoising.  MEND enables low intensity peaks to be detected, thus providing additional useful information for sample analysis.The process of denoising, performed in the chromatographic time domain, does not distort peak shapes in the m/z domain, allowing accurate determination of MS peak centroids, including low intensity peaks.MEND has been successfully applied to denoising of LC-MS data sets generated by different types of instruments, e.g. LC-MALDI-TOF-MS (AB 4700 Proteomic Analyzer, 2 kHz MALDI-TOF instrument made in lab), LC-MALDI-QqTOF-MS (Manitoba/Sciex prototype), LC-ESI-TOF-MS (Mariner, AB and LC/MSD-TOF, Agilent).Recently MEND has been used for denoising and quantitation of data sets from ion trap and FT mass spectrometers as well (LCQ, LTQ, LTQ-FT, Thermo).MEND has been shown to suppress chemical and random noise, base line fluctuations, as well as filter out false peaks originating from the matrix (MALDI) or mobile phase (ESI).

Selection of MS/MS precursor ions by PRESEL

Using LC-MALDI MS/MS, a new algorithm (PRESEL) has been developed for selection of precursor ions and determination of their optimum positions on the MALDI plate for MS/MS acquisitions in order to maximize the number of peptide identifications.In LC-MALDI MS analysis of complex peptide mixtures, the number of coeluting peptides can be frequently greater than the maximum number of MS/MS spectra that can be acquired from a given time interval of the chromatogram. For a given MALDI spot, the more MS/MS acquisitions (i.e. laser shots) that occur before a given precursor is analyzed, the lower will be the amount of material available at that spot.This will result in a decreased MS/MS fragment ion intensity and a subsequent lowered probability of peptide identification. PRESEL increases the number of identified peptides in an LC-MALDI MS analysis by redistributing positions for MS/MS acquisitions.The main criterion for selection of optimum precursor ion positions is effective intensity – the MS intensity of the precursor ion corrected for the decay in the MS/MS fragment ion intensity as a result of prior MS/MS acquisitions from the given spot.Comparison of the results of the LC-MALDI MS/MS analysis of strong cation exchange fractions of the tryptic digest of yeast lysate with and without processing by PRESEL shows a 40% gain in the number of identified peptides due to redistribution by the PRESEL algorithm.

Denoising of MS/MS spectra by wavelets

Wavelet denoising is known to be efficient for analysis of signals presented as a succession of peaks, e.g. chromatograms and MS spectra, In our laboratory an algorithm for wavelet denoising of MS/MS spectra was developed.The optimization of wavelet denoising was performed with the subset of 200 MS/MS spectra by varying the wavelet base functions, order and the level of decomposition.The Symlet class of wavelet bases was used for denoising because of its nearly symmetrical shape similar to MS/MS peaks.In particular, the sym8 function was used for denoising with the level of decomposition equal 5 and soft thresholding according to applied at each level. An example of denoised spectra is presented below. It can be seen that denoising recovered almost the complete y-ion series without compromising mass accuracy or mass resolution resulting in identification of peptide that was not identified without MS/MS denoising.

References

1. Victor P.Andreev, Tomas Rejtar, Hsuan-Shen Chen, Eugene V. Moskovets, Alexandr R. Ivanov, and Barry L. Karger. A Universal Denoising and Peak Picking Algorithm for LC-MS Based on Matched Filtration in the Chromatographic Time Domain. Anal. Chem. 2003, 75, 6314-6326.

2. Victor P.Andreev, Tomas Rejtar, Hsuan-Shen Chen, Eugene V. Moskovets, Alexandr R. Ivanov, and Barry L. Karger. A New Algorithm for Minimizing Chemical Noise in LC-MS: Matched Filtration with Experimental Noise Determination (MEND). ASMS 2003.