MS2DB: An Algorithmic Approach to Determine Disulfide Linkage Patterns

Abstract

Determining the number and location of disulfide bonds within a protein provide valuable insight into the protein's three-dimensional structure. Purely computational methods that predict the bonded cysteine pairings given a protein's primary structure have limitations in both prediction correctness and the number of bonds that can be predicted. Our approach utilizes tandem mass spectrometric (MS/MS) experimental procedures that produce spectra of protein fragments joined by a disulfide bond. This allows the limitations in correctness and scaling to be overcome. The algorithmic problem then becomes how to match a theoretical mass space of all possible bonded fragments against the MS/MS data. In our algorithm, which we call the indexed approach, the regions of the mass space that contain masses comparable to the MS/MS spectrum masses are located before the match is determined. We have developed a software package, MS2DB, which implements this approach. A performance study shows that the indexed approach determines disulfide bond linkage patterns both correctly and efficiently