Sifting through the tandem MS trash for proteins?

Review: Dynamic Spectrum Quality Assessment and Iterative Computational Analysis of Shotgun Proteomic Data: Toward More Efficient Identification of Post-translational Modifications, Sequence Polymorphisms, and Novel Peptides. By Alexey I. Nesvizhskii et al.
Molecular & Cellular Proteomics 5:652-670, 2006

The Aebersold group continues to produce great new tools for proteomic data analysis. Presented is a method to comb MS/MS data for good spectra which were not matched with a peptide on your first search. Unique to this technique is the notion that what constitutes a high quality spectrum can be learned from the analyzed data itself, i.e., without relying on a training data. The database search assignments that are made in the first pass at data analysis are generally based on high quality MS/MS spectra, these are used to identify high quality spectra that were not matched. The method is based on a Linear Discriminant Analysis of a set of spectra quality measurements.

A nice aspect of this paper is that the authors did a good job of illustrating how their nifty new data analysis tool could result in biological insight. One drawback is that a lot of computation is required to squeeze a little more information out of your data.

This software is named QualScore and can be found at: http://cvs.sourceforge.net/viewcvs.py/sashimi/