No calculated concentration values randomly Serge  2020-02-13
 

Hi,
I use the software version 20.1 (64 bit).
I use two transition (Analyte and internal standard). I generate calibration curve using peak area ratio but for some samples, they didn't calculate the concentration. Those samples with uncalculated concentration had appropriate peak ratio value.
In attachment, you have different print screen of the method.
I'm waiting your feedback
Serge

 
 
Nick Shulman responded:  2020-02-13
The warning that you are seeing on the calibration curve window is "The selected replicate has missing or truncated transitions". This is probably the cause of the behavior you are getting, and it means that Skyline does not think it has a valid peak area for one or more transitions in the current replicate.

For the calibration curve, Skyline wants to make sure that it is doing the quantification calculation using the same set of transitions across all of the replicates. The error message is telling you that the replicate that you have selected is missing values for some of the transitions that do have values in the external standard replicates.

If you would like, you can send us your Skyline document and I could tell you more specifically what is going on.

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.

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

-- Nick
 
Serge responded:  2020-02-13
Hi Nick,
I join the Skyline file with data. Let me know if you need other document.

Serge
 
Nick Shulman responded:  2020-02-13
Serge,

You did not actually attach the .zip file, so I cannot see what your chromatograms look like.
I can tell from the .sky file that you attached that Skyline says that the peak that it chose for "Analyte" in Replicate 15 was "truncated".

Truncated means that the chosen peak appears to be cut off because either the start or end boundary of the peak coincides with the start or end of the selected or extracted chromatogram. In general, you cannot trust the area of a truncated peak because you cannot be sure how much of the peak area is missing.
-- Nick
 
Serge responded:  2020-02-14
Hi Nick,

You're right, Skyline didn't calculate concentration because it detects peak as ''Truncated''. For our application, Skyline qualified the peak as ''Truncated'' but the peak area is fine. Can we change the criteria for truncated peak detection?

Serge
 
Brendan MacLean responded:  2020-02-15

Hi Serge,
You can always "rescue" a peak classified as "truncated" by moving the integration boundary at the edge one point inward. That is, the current criteria Skyline uses to classify a peak as truncated are:

  1. An integration boundary is placed at one extreme end of the available chromatogram (i.e. first or last point in the chromatogram)
  2. The integration boundary at the chromatogram extreme is more than 1% of the peak height higher than the other integration boundary.

I agree that #2 is too stringent and should probably be raised to something like 10%. I am the one who wrote this algorithm, but I have since seen peaks I also felt were not really truncated that get classified as truncated, and 1% of the peak height now seems to me like it could easily just be caused by noise variation. So, probably we will change this.

For now, if you want to manually correct this, you can simply move the integration boundary one point in from the edge. Skyline also gives you a number of tools to find which peaks have been classified as truncated, including an advanced option under Edit > Find and the Document Grid.

This is covered in great detail starting on page 11 of the Processing Grouped Study Data tutorial:

https://skyline.ms/_webdav/home/software/Skyline/@files/tutorials/GroupedStudies1-3_1.pdf#page=11

Good luck. Thanks for reporting the issue. Look for a change to a higher percentage of the peak height in the future.

--Brendan

 
Serge responded:  2020-02-17

Hi Brendan,

Thank you. Your comment helps me to fix this issue. I'll test your further version with higher percentage. Let me know, when it'll be available.

Serge