COMPUTER-AIDED METHODS FOR IDENTIFYING POST-TRANSLATIONAL MODIFICATIONS IN LC-MS OF PROTEIN DIGESTS.

J. Alex Taylor, Kenneth A. Walsh and Richard S. Johnson (1). Department of Biochemistry, University of Washington, Seattle, WA 98195, USA ; (1) Immunex Corp., 51 University St., Seattle, WA 98101, USA

Abstract

A program, Sherpa, has been developed which provides a battery of searches to allow an investigator to interpret more efficiently LC/MS and MS/MS data from protein digests by facilitating first a distinction between signal and noise, and then comparisons of observed data with peptides predicted from proteins of know sequence. However, as virtually all proteins are post-translationally modified, more is needed than just a simple matching of the data to the protein sequence. Sherpa contains specific search functions for phosphopeptides and glycopeptides. Both searches utilize ion groups and chromatographic tolerances to improve their accuracy. A built-in rating system for the results of searches draws attention to the most likely interpretations for further evaluation without excluding more remote possibilities from consideration. The flexible, nonlinear structure of Sherpa and its simple user interface allows hypotheses to be quickly tested under different settings and tolerances. As a result, the amount of manual interpretation needed is greatly reduced and focused.

LC/MS Interpretation


Fig. 1

The total ion chromatogram (TIC) after subtraction of background (upper), and the corresponding LC/MS contour plot from a tryptic digest of phosphorylated MBP.




Legend to Table 1
Summary of the results of comparing the processed LC/MS data to peptides predicted for a tryptic digest of MBP. Using an intensity threshold of 15000 cpm and minimum intensity of 24000 cpm for a peak top, 312 ions were observed. Ion groups were sought within a charge-state range of +1 to +5. A potential charge-state range of +1 to +2 was used for single ion peptide matching. Peptides were matched using a tolerance of +0.75 Da. with theoretical calculations of mass using a crossover point at 1500 Da. from monoisotopic to average mass.
a) The most likely peptide identifications are those that match an ion group composed of two or more ions of consecutive charge state. "Observed group mass" is an intensity-weighted average mass calculated from the masses of each observed ion in the group. "Group apexscan" is the intensity-weighted average scan at the maximal elution of the peptide. "Charge-states" refer to the positive charge forms found in that group. "Calculated peptide mass" is that predicted for a hypothetical tryptic peptide that is found to be within tolerance of the mass of the observed ion group. "Residues" and "Sequence" locate the peptide within MBP (immediately adjacent residues are also indicated). In the right-hand column, a closed circle indicates that MS/MS experiments confirmed the proposed match. An open circle indicates that MS/MS contradicted the proposed match.
b) A less likely list of peptide matches relies on single ion matches to peptides. In this example, an upper charge-state limit of +2 was specified for single ions. Changing the limit to +5 (not illustrated) increased the number of ions in this category from 26 to 41.

Poster Page 1 Poster Page 2 Poster Page 3
Sherpa Home Page Univ. of Washington
Last Updated April 12, 1996
J. Alex Taylor (jataylor@hairyfatguy.com)