Correlation and convolution analysis of peptide mass spectra.

As proteomics continues to establish itself as an effective postgenomic research tool, there is an increasingly urgent need for efficient, automated analysis techniques capable of effectively dealing with the vast amounts of data generated via mass spectrometry. Wholesale analysis packages, often used to deal with these enormous amounts of data, may benefit from supplementary, targeted analyses as current research begins to emphasize posttranscriptional/translational protein modifications, protein truncations, and poorly characterized mutations. We demonstrate the application of a new analysis technique based on mathematical correlation that is computationally efficient and robust against different instruments, noise levels, and experimental conditions. We have previously shown that this technique is able to extract pertinent mass shift signals from MS data, corresponding to the neutral loss of a modification from a peptide, e.g., a loss of 79.97 Th from phosphorylated tyrosine. Here we show that an extension of this method is applicable to MS and MS/MS data in general, allowing visualization of ions that produce a particular mass shift signal, be it from differential stable isotope labeling, overlap of fragment ions in a series, or ions that produce a neutral loss. The application of this method allows the researcher to discover individual features, such as the presence of specific modified or isotopically labeled peptides, to eliminate overlapping fragment ion series, and to localize specific sites of modification.