Peak area integration error

support
Peak area integration error joshua shipman  2024-07-19 11:38
 

I am encountering an error where single scan frames will produce peak areas that measure much higher than they should. In the attached file "skyline error" the three samples are replicates, you can see that scan 3 has a much higher measured area. When you shift the scan over a single frame in either direction like in file "skyline error corrected" the peak area changes by ~100% and the replicates will have similar areas as expected.

 
 
Nick Shulman responded:  2024-07-19 11:42
Could you send us a couple of .sky.zip files instead of the .sky files that you have attached?

In Skyline you can use the menu item:
File > Share
to create a .sky.zip file which will contain the .sky file that you attached as well as supporting files including extracted chromatograms.

Files which are less than 50MB can be attached to these support requests.
You can always upload larger files here:
https://skyline.ms/files.url
-- Nick
 
joshua shipman responded:  2024-07-19 12:01
Thanks Nick,

File is uploaded as "07192024 JTS.sky.zip"

I didn't upload the corrected one but just note what happens to the peak areas when you shift the beginning of the retention time window around for the third sample.

Josh
 
Nick Shulman responded:  2024-07-19 14:48
Thank you for reporting this bug.

When you adjust the start time of that particular peak it has a huge impact on the reported area.
The reason for this is that Skyline is looking at the time interval between the point at left edge of the peak, and the neighbor to its immediate right, and that number is being used to convert between the sum of heights and the reported area.

Skyline seems to have a potential to miscalculate areas when number of points in the chromatograms approaches or exceeds 1000 per minute.
Skyline usually resamples the chromatogram points in the time dimension so that when it calculates the area under the curve the points are evenly spaced. However, Skyline does not do this resampling if it sees that the variation in these time intervals is less than 0.001 minutes.

When we wrote this code back in 2009 we must have been assuming that the number of chromatogram points would never even be close to 1000 per minute.

I will try to fix this in an upcoming release of Skyline-daily.
-- Nick
 
joshua shipman responded:  2024-07-22 06:10
Thank Nick, just some background this is CZE-MS data collected on a ZipChip, which is why the peaks are so narrow (typical peptide peak with is <2 seconds).