Re: ms/ms fragment in TurboSEQUEST searching
Steven,
Sequest assumes that the ion type is y", even though it is labeled as y, so I don't think that is your problem. Have you constructed your own database and applied all modifications that you expect (e.g., alkylated cysteines, etc)? If you have, then I'd look at the masses that I'm getting and compare them with what I would expect. What I'm suggesting is: (1) make a list of all digest masses that you expect and (2) compare this list with the list of masses determined from the Sequest search. This is a fairly tedious process, but at least it will tell you what may be going wrong. If Sequest gets the wrong mass, then the ms/ms search will not work, since it finds all peptides matching the determined mass before performing the ms/ms correlation. Are you using zoom scan? If so the calculated mass from the sequest search is pretty good, assuming the charge was determined properly.
I usually use a peptide mass tolerance of 3 Da. This mass tolerance only has an effect on selecting candidate peptides prior to cross-correlation. It does not have any effect on the tolerance of the fragment ion masses. I usually like to use average mass for parent ions and monoisotopic masses for fragment ions. The reason for this is that the parent ions are often multiply-charged and their m/z values are usually better compared using average mass. Most fragment ions are usually singly charged, so using monoisotopic mass for these is usually best. You set these options in the TurboSequest dialogue under the Ions and Modifications tab. In any event, if you have your peptide mass tolerance set to 3 Da, then setting parent mass type to average or monoisotopic will not make much of a difference.
©Copyright 2000-2008 Novatia LLC
Disclaimer and terms of usage
e-mail: info@enovatia.com
|
