Specialized high-res analysis of PRM data

support
Specialized high-res analysis of PRM data Tomas Vaisar  2023-05-05 07:18
 

Hi Skyline team,
I have a specialized analysis for in vivo protein turnover kinetics data analysis. In a nutshell this is for D3-Leu in vivo enrichment in people, where I run the PRM with precursor selection on average of the H3-Leu/D3-Leu precursor mass (and wide selection window) and very-high res MS/MS (240k). I want to be picking up D3-Leu fragments selectively from A+3 of the H3-Leu fragment. I tried to upload this data into Skyline, but it does not recognize the precursor correctly (because it is average of the H3-Leu and D3-Leu version of the peptide - e.g. for peptide IGVELTGR 422.748 and 424.2574 the precursor in mass spec was set to 423.5080). Is there a way to trick Skyline to upload this data? I believe that the resolution for the MSMS should be easy to set.
Thanks,
Tomas

 
 
Brian Pratt responded:  2023-05-05 13:53

Hi Tomas,

I'm not sure I understand completely, but it sounds like you just need to open up the MS1 tolerance until the machine value falls within the range of both target precursor m/z? ? If you can provide the various files involved we can probably figure out a solution.

Thanks for using the Skyline support board,

Brian Pratt

 
Tomas Vaisar responded:  2023-05-05 14:05

Thanks Brian,
There are two issues - 1. the precursor mass does not match either the H3-Leu peptide mass nor the D3-Leu peptide mass (it is average of the two, and b) the precursor selection window was 4 Da.
I thought I could open the precursor window and upload to Skyline, but it did not work.

Attached are the raw file, method file and a Skyline file with a subset of peptides (not all precursors acquired are in it).

Thanks,
Tomas

 
Tomas Vaisar responded:  2023-05-05 14:06

Thanks Brian,
There are two issues - 1. the precursor mass does not match either the H3-Leu peptide mass nor the D3-Leu peptide mass (it is average of the two, and b) the precursor selection window was 4 Da.
I thought I could open the precursor window and upload to Skyline, but it did not work.

Attached are the raw file, method file and a Skyline file with a subset of peptides (not all precursors acquired are in it).
I cannot attach the raw file - where do you want me to put it?

Thanks,
Tomas

 
Brian Pratt responded:  2023-05-05 14:09

You can drop it at http://skyline.ms/files.url

 
Tomas Vaisar responded:  2023-05-05 14:17

File 230421S211KinTest_C12T12_PRMHiRes4.raw
Just put it there.
Thanks,
Tomas

 
Brian Pratt responded:  2023-05-05 14:40

Do you have an expected retention time for IGVELTGR? Looking at the .raw in SeeMS it looks like the desired fragments just aren't there?

 
Tomas Vaisar responded:  2023-05-05 14:58

IGVELTGR peptide has retention time 15.23 min. It has decent intensity. Fragment ions to look at are 674.383, 575.315, and 446.272 for the H3-Leu version, and 677.402, 578.333, and 449.291.

Tomas

 
Brian Pratt responded:  2023-05-05 16:53

The trick would be to raise Transition Settings > Instrument > Method Match Tolerance m/z enough to cover that difference. The problem is that we enforce a limit of 0.6 on that setting, and the difference is something like .8.

I can make this work if I set Transition Settings > Instrument > Method Match Tolerance m/z to 0.8 - but I can only do that because I temporarily altered the source code.

I'll have to ask around here to understand why we have that setting limited that way.

 
Tomas Vaisar responded:  2023-05-05 17:03

Thanks,
That's exactly where I went first and found the same thing.
I do not think there is any reason to keep the limit on this parameter other than protect users from their own folly.
it would be great to be able to go higher than 0.6 (or even 0.8).
Thanks,
Tomas

 
Brendan MacLean responded:  2023-05-08 09:01

What we now do for the "PRM" acquisition method is achievable with a "DIA" acquisition method. In fact, for a long time, Skyline didn't have this PRM option, but only "Targeted" which allowed only 1 precursor m/z to extract from the MS/MS spectra. The target precursor m/z closest to the method precursor m/z always won. This turned out not to be such a great strategy, and we ended up telling people to use a DIA setting to achieve what is now PRM.

The "PRM" method is just shorthand for this method where the isolation range is taken from the Method match tolerance. So, you are right to look there, but when you reach that limit, you are obviously doing something a bit experimental, and you should just switch to DIA, which allows an arbitrary isolation range.

I think your data comes out looking pretty good if I use Acquisition method = "DIA", Isolation scheme = "Results only". If on the other hand, you really want to restrict precursor matching to something smaller than the instrument isolation range, you could add a margin (there is "Results only (0.5 margin)" by default) or you can play around with defining your own margin or even a fixed range, like 1.6 m/z which seems to be what you were trying to achieve above.

Hope this helps. I think we have plenty of flexibility in this area. It just requires switching your thinking from PRM to DIA with a Results only isolation scheme, which is all that PRM really is in its implementation (DIA with Results only and a Fixed isolation width matching 2x the Method match tolerance).

--Brendan

 
Brendan MacLean responded:  2023-05-08 09:10

Oops. Here is the document with settings set up as I described and the data imported with nice chromatogram peaks.

 
Tomas Vaisar responded:  2023-05-08 09:25

Brendan,
Of course!!! This makes perfect sense and I should have thought about it. Yeah, it was the naming "PRM" that blinded me when in fact it is version of DIA. I will run through the data and will let you know if there are any issues.

Really appreciate your help!
Tomas

 
Tomas Vaisar responded:  2023-05-08 09:38

All works great. BTW - there is no way to normalize peak areas (Replicate comparison) to "light" rather than "Heavy"?
I am looking at extent of heavy label incorporation.

Tomas