IMS Filter Error with version 20.2.0.343 Juan C. Rojas E.  2021-01-25
 

Dear support,

I am having some issues with applying the IMS filter windows generated from the in-built predictor for DDA data with IMS separation. The data was generated with a Synapt G2-Si.

The steps I usually follow are:

  • Load DDA data
  • Filter reliable peptides
  • Use IMS Predictor
  • Re-import results to apply IMS filter
  • (New due to change in settings) Set Window type as Resolving Power and set value of "15"; this was the optimal value found for my data in the past. BTW, why was this setting changed out of the library generation window? Shouldn't the resolving power settings be set prior to generating the IMS library?

This worked fine in version 20.1.0.155. However, I recently updated to the newer version 20.2.0.343 and every time I try to do the re-import I get an error message "An item with the same key has already been added.". Additional text (in German) attached.

I also tried removing the file, save, set IMS library created previously, set IMS filter settings, and then import the file again. This time the file loads, but no IMS filter is applied.

Can you give me some insight for this problem? What else could I share to help?

Sincerely,
JC

 
 
Juan C. Rojas E. responded:  2021-01-25

I forgot to add that I also tested with Skyline Daily (20.2.1.384) and I am having the same error messages.

Thanks for the time and support.
JC

 
Nick Shulman responded:  2021-01-25
JC,

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 your ion mobility library.

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

I am not the expert on ion mobility libraries, but it looks like the bug has something to do with the same peptide or molecule having more than one listing in the library. This bug looks like it might have been introduced in July 2020.
-- Nick
 
Juan C. Rojas E. responded:  2021-01-25
Hi Nick,

I have uploaded the file to the link you sent with the following name:

ARP-HumanPl_2000ug_pool_t2_reanalysis_Transfer_cal.sky

Well, the data is for modified peptides where for some cases I have identified isomers with distinct elution times. Therefore, I created a copy of the peptide in the document to distinguish them.

In the past I wondered how the predictor handled this since there usually was just one entry per peptide sequence and there was no way of distinguishing these isomers. But I guess it just ignored one of them.

- JC
 
Juan C. Rojas E. responded:  2021-01-25
Nick,

I went ahead and deleted all the isomer "copies" and attempted the initial workflow. Without the "copies" the IMS filter was successfully applied as before.

Although this is a functional workaround, it would be a nice feature request allowing the consideration of these isomers in future releases.

Thanks for the info!

Sincerely,
JC
 
Brian Pratt responded:  2021-01-25
Hi Juan,

Can we have one of the mass spec data files as well, so we can reproduce your steps?

Thanks,

Brian Pratt
 
Juan C. Rojas E. responded:  2021-01-25
Hi Brian,

I just uploaded the file: "S4_20_39_2.raw.zip". With an excess of precaution, I would like to ask if you could please delete it from the repository as soon as you get it?

Note: not all peptides in the document are supported by the MSMS of that file.

- JC
 
Brian Pratt responded:  2021-01-25
I would think that two molecules with different elution characteristics would also have different ion mobility characteristics.

What changed here is that Skyline moved to a library model for ion mobility - as with spectral libraries it's expected that there's a one to one relationship between a molecule and its metadata (ion mobility, in this case). Here you're dealing with two different molecules but not distinguishing them in any way. I'm surprised that this is the only problem you've encountered in Skyline with that approach, actually - although I understand why you've gone this route.

I'm really not sure what the right move is here - I think assigning them both the same ion mobility would be a mistake. With small molecules it's easy to just rename one of them, but with peptides the sequence IS the name so it's not really possible to distinguish them.

Thoughts?

Thanks

Brian
 
Brian Pratt responded:  2021-01-25
Thanks, I've downloaded and deleted the data file.
 
Juan C. Rojas E. responded:  2021-01-25
I have wondered the same about the ion mobility, but I have not developed more on this idea since I had a lot of other peptides to deal with. And if there is a distinction I am not sure that the resolving power in the IMS dimension would be able to resolve them fully with the MS method I have been using. Maybe fragments carrying the PTM would carry larger differences with their ion mobility. However, this is not a feature that is possible in Skyline with the current IMS filter since precursors and fragments carry the same IMS due to the fragmentation happening after IMS.

I will try to look into the differences in IM of the precursors.

The peptide Identification approach I used also can't distinguish these species since they carry the same fragment (although at slightly different relative abundances).

The only way I could think of distinguishing these peptides in an automatic way is associating the RT to the ion mobility. However, that comes with implications to stable RT or some sort of RT calibration done a priori. Otherwise allowing manual annotation of these peptide cases by identifying duplicate sequences with same charge states and forcing the user to write down the estimated center of the IM filter window by checking the MS1 traces.

Sincerely,
JC
 
Brian Pratt responded:  2021-01-25
Yes, I'd thought about associating RT as well but that does add a lot of complexity.

>> forcing the user to write down the estimated center of the IM filter window by checking the MS1 traces
This can already be done, if needed, by setting the Explicit Ion Mobility value for each instance of a precursor in the Document Grid. But it's not much fun.

I'll change Skyline to assume that all instances of a precursor are going to have the same IM value regardless of the chromatography.

Best regards,

Brian
 
Juan C. Rojas E. responded:  2021-06-21
Dear all,

I have to revisit this topic because I am encountering some issues with the IMS filters again.

Currently using: Skyline-daily 21.1.1.160

I have found that I tend to approximate better the offset by using a systematic offset centered at the median of all the estimated offsets, either by using the results in Skyline or from a search engine that uses IMS filtering (e.g. PLGS). So I reintegrate my data compatible for IMS filtration with the "fixed" IMS tables and tends to work pretty well. However, every now and then the precursor ion's ion mobility center is still incorrectly assigned, requiring some manual "fixes" as I tried in the example attached. I have tried to re-import my results multiple times, but for whatever reason the changes I made in the "Edit Ion Mobility Library" are not being recognized. Any ideas why? Suggestions?

As always, thanks for the time and help.
Sincerely,
JC

PS: I have removed any duplicate peptides to avoid the issue outlined in previous entries of this thread. Out of curiosity, was "I'll change Skyline to assume that all instances of a precursor are going to have the same IM value regardless of the chromatography" implemented to the most recent versions of Skyline?
 
Brian Pratt responded:  2021-06-21
It's probably just the effect described here: https://skyline.ms/wiki/home/software/Skyline/page.view?name=Ion%20Mobility%20Separation%20%28IMS%2C%20or%20Drift%20Time%29%20Data

"... It's important to understand that the CCS value takes priority: different raw data files may contain different CCS->mobility calibrations, so the actual ion mobility filter value for a chromatogram extraction is always derived from the CCS value when available. Thus, if you want to experiment with adjusting ion mobility values for chromatogram extraction, it's import to either adjust CCS rather than ion mobility, or to clear the CCS setting so that your adjusted ion mobility value is the one that gets used. "

It should be the case in the current Skyline that two different instances of a peptide precursor will draw the same ion mobility value from the ion mobility library.

Best Regards,

Brian
 
Juan C. Rojas E. responded:  2021-06-22
Indeed it was the priority for CCS you pointed out. I deleted all the CCS values since we are not interested in reporting an accurate CCS estimation for what we need with the current project. Re-imported and all the manual changes to the ion mobility (center) were accepted.

Thank you very much for the clarification.

Sincerely,
JC