• Sign In
  1. Requests

Peptide settings (# per protein) not working with decoys

support

View Request

view list print
Peptide settings (# per protein) not working with decoys liyan_c  2016-11-16 05:30
 
Hi Skyline team,

I am using Skyline 3.6 (administrator install) in Win 7 pro for DIA data acquired on Thermo Q-exactive HF. Here is what I did today:

Setup a document with maximum of 5 peptides/protein under library options:
Added equal number of decoy peptides as my target list (~21.000)
Imported DIA files
Trained decoy model using
Re-integrated peptides with Q value > 0.01

After this was done, I found that Skyline had added all library peptides into the targets list, even non-integrated ones. I have checked that "integrate all" was not selected.

Here are some things I did to refresh the peptide settings:
Closed Skyline, reopened the document and changed # of peptides/protein to 6, then back to 5
Re-scoring data in place
Remove empty peptides under "refine"
 
The extra peptides are still stuck in the document (please refer to screenshot). The only settings that still work for this document are refinement based on RT regression, dotp and idotp.

Am I doing something incorrectly?
 
 
Brendan MacLean responded:  2016-11-16 05:51
Maybe the best way to deal with this would be to set up a report in the Skyline Document Grid that shows only peptides matching your filter criteria, e.g. q value > 0.01, copy all the peptide sequences or modified peptide sequences, and paste them into the Edit > Refine > Accept Peptides form, and then click the OK button. This will remove everything not in your set. It does not matter if your set is repetitive.

This methodology is covered step-by-step in the Skyline Tutorial Webinar #8: DDA to Targeted: Differential Statistics with Skyline (starting around minute 40 of the video):

https://skyline.ms/webinar08.url

In this case, the report is used to identify peptides with significant differential changes. But, the principle is the same for any set of peptides you wish to keep in the document, no matter how you determine this.

I agree that we should probably add something more specialized for handling this case after peak picking with an mProphet model. But, until we implement that, you can always use this method of selecting peptides to keep in your document and removing all others.

Hope this helps. Thanks for posting your question to the support board.

--Brendan
 
liyan_c responded:  2016-11-20 18:34
Hi Brendan,

Thanks for the workaround, my document is fixed now.