Decoys not showing up with dotp values from Prosit library cm748  2022-03-07 04:04
 

I've been trying to create an mProphet peak scoring model to pick my peaks but I'm running into a bit of a problem including the library intensity dot product field... I have generated my document with decoys (with random mass shift) and acquired data for the for the targets and decoys (with RepLiCal iRT std for RT prediction), then built and implemented a Prosit library to get the dotp values. However, when I implement the library the decoys do not have a dotp value (the targets do), even though in the view library match window prosit is predicting spectra for them.
I have included a screenshot, where you can see in the targets window for the peptide LHIS[...] that there is a dotp for the target but not for the decoy, even though prosit is showing a prediction for the decoy in the library match window.
Any ideas what could be causing this?

Thanks,
Colleen Maxwell
 
 
Nick Shulman responded:  2022-03-07 06:58
Can you send us your Skyline document?

In Skyline you can use the menu item:
File > Share
to create a .zip file containing your Skyline document and supporting files including extracted chromatograms and spectral libraries.

If that .zip file is less than 50MB you can attach it to this support request. You can upload larger files here:
https://skyline.ms/files.url

It looks like there must be a bug in Skyline, and seeing your Skyline document will be really helpful in terms of figuring out what that bug is.
-- Nick
 
cm748 responded:  2022-03-07 08:00
Hi Nick,
Thanks for the quick response,
I've attached the .zip file here containing the document, the MS files and the spec lib I generated with Prosit

Thanks,
Colleen
 
Nick Shulman responded:  2022-03-07 09:04
I see that you used "Random mass shift" as your decoy generation method.
It looks like if you had used one of the other methods such as "Shuffle sequence" or "Reverse sequence" then the decoy peptides would have remembered what sequence was used to generate them, and they would display that source peptide's spectrum information. For some reason, it's not doing the correct thing for random mass shift. I am not sure whether we meant to do it this way, or if it's a bug.

I will look into this more and figure out how it's supposed to work.
-- Nick
 
cm748 responded:  2022-03-07 12:02
Thank you Nick, I will try using the other methods of decoy generation and see if that works for me
Colleen
 
cm748 responded:  2022-03-07 13:21
Using reverse sequence instead of random mass shift has worked - I now have dotp values for both the targets and the decoys. Thanks for the help!
Colleen
 
Nick Shulman responded:  2022-03-08 11:45
Colleen,

I have looked into this a little bit more, and I think the behavior that you are seeing has nothing to do with decoy mass shift etc.
The thing that actually fixed this problem for you was just doing "Refine > Add Decoys". Any option that you might have chosen there would have resulted in the decoys having spectra associated with them.

I think what happened is that you first added decoys to your document when your document had no spectral libraries in it. When you added your prosit library to the document, Skyline updated the library information for the real peptides in your document, but that information was never updated for the decoy peptides. This seems to be the intentional behavior in Skyline. Whenever you make a settings change which could potentially affect the transitions that Skyline will choose for your peptides, you should tell Skyline to generate decoys again by going to:
Refine > Add Decoys

-- Nick