Intensity difference between skyline and xcalibur

support
Intensity difference between skyline and xcalibur dtan  2018-10-23
 

Dear Skyline team,

I'm using skyline to extract the intensity of fragment ions from PRM data for small molecules. However, I found that the intensity extracted by skyline was much higher than that extracted by Xcalibur (~20E6 vs 4.99E6). What should I do about this?

Thanks,
Dan

 
 
Brendan MacLean responded:  2018-10-23

Hi Dan,
I recommend that you convince yourself that the two peak areas are highly linearly correlated and then not worry that these unitless values don't match exactly. If you want, you could use a derived linear equation to convert between the two, as we did for the DIA tools assessed in this paper, where also raw peak areas differed widely, but the tools were found to be highly correlated:

https://www.ncbi.nlm.nih.gov/pubmed/27701404

We have no visibility into how Xcalibur calculates its peak areas, but the algorithm used for ours is open source and described here:

https://skyline.ms/wiki/home/software/Skyline/page.view?name=tip_peak_calc

We understand that the two do not match, but there is ample evidence that Skyline peak areas are accepted in the field.

Thanks for posting to the Skyline support board.

--Brendan

 
Nick Shulman responded:  2018-10-23
Can you send us your Skyline document and .raw file? I will be able to tell you how Skyline got the numbers that it did.

I am not sure whether we ever tried to make it so that the intensities reported by Skyline would be identical to the ones in XCalibur, or whether we just guaranteed that they would be proportional.

If you click on a chromatogram in Skyline, it brings up the Full Scan window, where Skyline will show you how wide of a channel in the m/z space Skyline summed across to get the chromatogram intensity. It might be that Skyline was summing across a wider window, but that would depend on the settings in "Transition Settings > Full Scan".

Anyway, if you send me your data I can take a look.

In Skyline, you can use the menu item:
File > Share > (complete)
to create a .zip file containing your Skyline document and supporting files including extracted chromatograms.

You can upload that .zip file here:
https://skyline.ms/files.url

You should also upload one of your .raw files.
-- Nick
 
dtan responded:  2018-10-24
Hi Nick and Brendan,

Thank you for your rapid reply and helpful information!

Nick, I uploaded my files as you suggested (20181024.zip). It'll be great if you can tell me how to set the parameters in "Transition Settings > Full Scan" to change the window.

Thank you,
Dan
 
Nick Shulman responded:  2018-10-24
Dan,

You can actually get the same numbers as XCalibur if you tell Skyline to centroid the data when extracting the chromatograms.
You can tell Skyline to centroid by changing the "Product Mass Analyzer" to "Centroid".
This setting is located at:
Settings > Transition Settings > Full Scan

When Skyline extracts chromatograms from profile data Skyline sums all of the profile intensities across a m/z range.
You can see that m/z range if you click on a chromatogram to bring up the Full Scan window in Skyline.
In the attached picture you can see that Skyline summed nine profile points between 674.5119 and 674.5253. That's how Skyline came up with the total sum of 19.3 million.

If you change the mass analyzer to "Centroid" then Skyline will use Thermo's centroiding algorithm so that there is only one point in the spectrum, and that point ends up with the intensity 4.99, which is the same as you see in XCalibur.

We do not know what Thermo's centroiding algorithm does, but when you tell Skyline to use it, it appears that Skyline will get the same numbers as XCalibur.
-- Nick
 
dtan responded:  2018-10-24
Got it! Thanks Nick!

Best,
Dan