Production of reliable MALDI spectra with quality threshold clustering of replicates

We present the first application of the quality threshold (QT) clustering algorithm to mass spectrometry (MS) data. The unique abilities of QT clustering to yield precision nodes that are commensurate with the mass measurement precision of the instrument are exploited to generate a consensus spectrum out of multiple replicate spectra. The spectral dot product and confidence intervals are used as a tool for evaluating the similarity and reproducibility between the consensus and replicates. The method is equally applicable to high and low resolution measurements. This paper demonstrates applications to linear spectra from a matrix assisted laser desorption ionization (MALDI) time of flight (TOF) instrument as well as peptide fragmentation data obtained from a TOF/TOF after unimolecular decomposition. The advantages of clustering to mitigate the inherent precision the shortcomings of MALDI data are discussed.