Separate MS/MS Filtering Settings for Light vs. Heavy Precursors

support
Separate MS/MS Filtering Settings for Light vs. Heavy Precursors n t wamsley  2021-07-19
 

Hello, I am working on a PRM method with stable-isotope-labeled internal standards. Technically it is "SureQuant" where the SIL peptides are monitored to trigger acquisition of endogenous peptides. MS2 scans for the heavy SIL peptides are carried out in an orbitrap at a resolving power of 7,500 at 200 m/z. However, MS2 scans for the endogenous/light peptides are carried out at a resolving power of 60,000 at 200 m/z. Right now I am using 7500 at 200 m/z for both the light and heavy peptides. However, I would like to enforce a separate and more stringent threshold for the endogenous MS2 scans that reflects the different instrument settings used in my method. I believe that on account of the thresholds being too low for the endogenous (in skyline), some of the endogenous transitions appear to have a lot of interference. At the moment is it possible to change this in skyline?

 
 
Nick Shulman responded:  2021-07-19
Title: Seperate MS/MS Filtering Settings for Light vs. Heavy Precursors
For most instruments, we recommend that you choose "Centroided" as the mass analyzers at "Settings > Transition Settings > Full Scan".

The resolution that you choose in the Transition Full Scan settings are not required to match what the instrument actually collected. The resolution (or mass accuracy) settings control how wide of a m/z channel Skyline sums across when extracting chromatograms from MS1 and MS2 spectra.

Skyline does not give you the ability to use different resolution settings for different Targets. It is conceivable that we should add this add a feature.

Can you send us some of your data? We might be able to come up with a different solution, or something else entirely might be going on which is unrelated to the resolution settings.

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.

You can upload that .zip file, and at least one of your .raw files here:
https://skyline.ms/files.url

-- Nick
 
n t wamsley responded:  2021-07-19
Nick,
     Thank you so much for your quick reply. I have uploaded the skyline document as instructed. We are using an Orbitrap Eclipse. I may switch to centroided as you said and play with the width of the m/z channel. I believe the trouble is that the MS2 scans for the SIL peptides are carried out at low fill times (10 ms) and low resolution, so I need skyline to sum over a wider m/z channel or else I will miss transitions due to poor mass accuracy. The SIL peptides are highly abundant, so I am less worried about interference. For the endogenous/light peptides, however, I am doing longer fill times (116 ms) and higher resolution, so a narrow m/z channel is appropriate for skyline. Since some of the endogenous peptides are low abundant and the matrix is complex, I believe setting a narrower m/z channel might reduce interference? It's clear from the chromatograms that for some of the low abundant peptides there are interfering signals for one or more transitions. I suppose that is unavoidable to some extent but could be minimized by tweaking the width of the m/z channel?
Best,
Nathan
 
Nick Shulman responded:  2021-07-19

Thank you for sending those files. Can you point me to a particular peptide (or several peptides) that we should be looking at?

By the way, there seems to have been a bug in Skyline and the file that you saved cannot be opened by Skyline. There are peptides in your document which use the modification "Label:13C(6)15N(4) (R)", but the modification in the Settings part of the document is "Label:13C(6)15N(4) (C-term R)".
Because of this, if you were to try to open the document that you uploaded, you would see the error:

The file contains an error on line 166 at column 12.
No modification named Label:13C(6)15N(4) (R) was found in this document. 

I am not sure exactly when this problem might have happened to this file. Have you been running into any problems where Skyline was refusing to open the document after you saved it?
I was able to repair the problem in the file using a text editor, and I am attaching a fixed version of that Skyline document to this support request.
-- Nick

 
n t wamsley responded:  2021-07-20

Nick,
Thanks for you help with the modification error. I have uploaded a screenshot of the light peptide chromatogram for the peptide, VAATSVITIVK, where the problem for that peptide is more obvious than in the .raw file I sent you. I am wondering if it is possible to calculate peak areas with one m/z channel width, change the width and then calculate again without changing the boundaries of integration. That way I could export two tables with the peak areas calculated differently. I would use one table to get peak areas for the heavy peptides and another to get peak areas for the light peptides. To do this must I re-score or re-import? It does not seem that re-scoring changes anything? If I must re-import, will the boundaries of integration be the same?
Best,
Nathan

 
Nick Shulman responded:  2021-07-20
Thank you for uploading that image.
By the way, files that are less than 50MB can be attached to this support request, which you might find more convenient than uploading.

If you change settings that affect the extraction width, the only way to see the results of that change would be by doing a "Reimport".
"Rescore" can be used for things that happen after chromatogram extraction, but which affect peak picking, such as changing the "Internal standard type" at Settings > Peptide Settings > Modifications.

When you do a reimport, or a rescore, any peak boundaries that you have manually adjusted will remain the same. For peaks that have not been user-modified, the boundaries might change if the new chromatogram data causes Skyline to pick a different peak.

For SureQuant methods, we recommend that you go to:
Settings > Transition Settings > Instrument
and check the checkbox that says "Triggered chromatogram extraction".
The major reason that you should do that is that Skyline will incorrectly estimate the background if the start and ends of the peak are truncated.
You can learn more about the triggered acquisition setting here:
https://skyline.ms/wiki/home/software/Skyline/page.view?name=TriggeredAcquisition

Attached is a screenshot showing that the background for the y9 chromatogram is very large (shaded rectangle at the bottom of the peak).

Here is a page that talks about how Skyline usually calculates background (when "triggered acquistion" is off):
https://skyline.ms/wiki/home/software/Skyline/page.view?name=tip_peak_calc

Also, if you click on points along a chromatogram, Skyline will bring up the full scan graph, which shows you a spectrum, and also shows the m/z channel that Skyline is summing across when extracting the chromatogram. In the attached image "Full Scan Graph", it looks like the y8 transition has interference from something that will not be able to be filtered out no matter how narrow you make the extraction width. When you have interference like this, it is recommended that you mark the transition as "Non quantitative", and then its area will not be included in things like "Total Area". You can mark a Transition as non-quantitative by right-clicking on it in the Targets tree. (or, you might want to simply delete that transition from your Skyline document)

One more thing: I noticed that at "Settings > Transition Settings > Filter" you have "Product Ion Selection > To" set to "last ion - 1". We normally recommend that you set that to "last ion". In a peptide with 8 amino acids in it, the "last ion" is y7 or b7. Some people choose "last ion - 1" because they incorrectly think the last ion would be y8 or b8, but that's not how Skyline counts things. There are a lot of example Skyline documents out there that are still using "last ion - 1" because we haven't been very good at spreading the word about what the recommended setting is.

-- Nick
 
n t wamsley responded:  2021-07-29
Nick,
Thanks for your feedback. "Triggered chromatogram extraction" makes a big difference. How does skyline chooses the peak boundaries (I'm using the default model)? What are reasons those boundaries might change as a result of narrowing the m/z channel under MS/MS filtering? Finally, if Skyline does not change manually adjusted peak boundaries after re-integration, then is there a way to "trick" skyline into treating all peak boundaries as manually adjusted before I do a reintegration? If that were possible I could get comparable peak areas for light and heavy peptides using different m/z channel widths.
Best,
Nathan