CE Optimization Method/Transition List Export does not seem to be working

support
CE Optimization Method/Transition List Export does not seem to be working Will Thompson  2017-08-30 06:35
 
Hi Brendan and Brian,

See attached document. When I attempt to export transition lists or methods with CE Optimization, the methods only seem to contain a single transition CE value. Additionally, and this may be a feature change request instead of a bug, but it appears that Skyline does not support transition-level CE optimization for small molecules; when you export a method/transition list the CE is the same for every precursor, even if you have inserted different CEs for different transitions using Edit->insert->transition list functionality. This is a little dangerous because it is 'hidden' that you're not using the correct CE for those transitions unless you dig into the exported transition list or method. CEs should be allowed to be optimized at the transition level for small molecules, just like they are for peptides.

I also noticed that the CE optimization import does not seem to work for small molecules (CE Optimize methods generated manually but tried to analyze in skyline as such).

ALso, there seems to be a little display bug when you right click on peak area view and choose Graph-> CV histogram, it permenantly stays checked, even with you try to go back to replicate comparison view. (screenshot attached)

Thanks, and let me know if you have any questions.

Will
 
 
Brendan MacLean responded:  2017-08-30 18:11
Hi Will,

So, it appears that the cause of the confusion is that you have Explicit Collision Energy set on all your transitions.

Not that we think it is right for Skyline to just silently export the single explicit CE for each transition when you have clearly indicated you want a transition list for CE optimization in the method/transition list export form, but that is what is causing this. You can easily work around the issue by doing the following:

  • View > Document Grid
  • Views > Small Molecule Transition List
  • Select all cells in the Explicit Collision Energy column (select the top cell and ctrl-shift-down-arrow)
  • Press the Delete key
After clearing these values, you should be able to export a collision energy optimization method/transition list with 11 CE values per transition centered around your "Small Molecules" linear equation (CE = precursor m/z * 0.0275 + 10.8).

That said, I am not entirely clear this is what you want. I could imagine that what you were intending was to have Skyline export a transition list with 11 CE values per transition centered around your explicit CE values, and that seems like it could now have the benefit of allowing second-round CE optimization. That if we enabled that, you could:

  • Perform first round CE optimization based on your linear equation
  • Save optimal CE information for every transition where an apex intensity was observed in the middle of the measured CE values
  • Set Explicit CE on any transition where apex intensity was observed on an extreme CE value
  • Perform second round CE optimization centered around the explicit CE values for only the transitions with an extreme apex
So, were you expecting to export 11 CE values centered around your linear equation, or 11 CE values centered around your explicit CE values? If the former, then follow the steps above as a workaround.

Then, consider what you think should happen when you click the OK button in the Export Transition List form for a document with Explicit CE values. Should we warn:

This document contains Explicit Collision Energy settings. Skyline will ignore these for collision energy optimization.

OK Cancel

Offer an option:

This document contains Explicit Collision Energy settings. How would you like to handle these?

(o) Optimize around the Explicit Collision Energy values when present ( ) Ignore them and optimize around the values predicted by "Small Molecules" predictor

OK Cancel

Something else? Just silently do one or the other option?

At least it is clear that just silently giving you only 1 entry per transition is not the right thing to do. At least you can still achieve the linear equation option through a bit more work.

Thanks for reporting this issue.

--Brendan

P.S. - Your CV Histogram issue is totally the new intended behavior. If you look closely you should still be able to find a tab in your layout for the CV Histogram, probably it is right next to the tab labeled "Replication Comparison" in the partial screen grab you posted in the PowerPoint slide (about 2/3 down on the left edge). It is now possible to show multiple graphs for peak area and retention time at the same time.

However, I can see that this is confusing. So, we are just going to remove the check marks in the menus and let them always just active the forms (as they already do). Check marks in the menus are more appropriate anyway when the menu item can be used to turn on and off the option. These are on-only menu items.

Thanks again for the feedback.

 
Brendan MacLean responded:  2017-09-10 14:28
Hi Will,
I made a Skyline-daily release which I hope just puts optimization CE values around your Explicit Collision Energy values, ignoring the linear equation specified in Transition Settings - Prediction. I am pretty sure this works at least for method export, as I had your document to test with, and I actually had to update one of our tests to continue passing with this change. So, I made sure it was testing that the right number of CE values were exported around the Explicit CE values.

What I am less sure about is whether importing the resulting data from a method like this and then either storing it in an optimization library or using it to export a fully optimized method correctly applies the maximum shift value to the Explicit CE values, and not the linear equation CE values, since I didn't have a good test for this. Maybe I will look into writing one next. Or you can send me a new document that proves this one way or the other.

I tried, and the result is in Skyline-daily 3.7.1.11357. Let me know how it works for you. Thanks again for providing the example.

--Brendan