Question of "Peak" Defination

support
Question of "Peak" Defination hzhu  2023-03-02 04:36
 

Dear Skyline Team,
This software helps me a lot in analyzing PRM and DIA data. Thank you.
And I am curious about how the software identify peaks in spectra. However I can't find the defination of that anywhere...
For example, this explanation (https://skyline.ms/wiki/home/software/Skyline/page.view?name=tip_peak_calc) makes me clear that how Skyline calculates peak areas, but why the time 14.4- 14.48 is defined as the start and end of the peak to calculate? I wonder if there is a document containing the algorithm about that, like treat the "10% intensity of climax" as the bottom.
Thanks again for your reply!

Best regards,
He Zhu

 
 
Nick Shulman responded:  2023-03-02 08:01
The first step of peak detection in Skyline involves taking the second derivative of the chromatogram, and finding the places where that second derivative crosses the X-axis.

You can see what the second derivative looks like by right-clicking on the chromatogram in Skyline and choosing "Transform > Second Derivative".
I would recommend using Skyline-daily to look at the second derivative chromatogram instead of Skyline 22.2, because in older versions of Skyline you would not see the parts of the second derivative chromatogram that were below the X-axis.

After Skyline has found pairs of places where the second derivative crosses the X-axis Skyline does some more work to widen the peak boundaries so that the peak boundaries encompass all of what a human would consider to be the peak (I am a little unsure how the widening of the peak boundaries algorithm works).

There is a bit more information about peak finding in section 3.3 of Dr Pino's "Skyline ecosystem" paper:
https://analyticalsciencejournals.onlinelibrary.wiley.com/doi/full/10.1002/mas.21540
-- Nick
 
hzhu responded:  2023-03-02 22:08
Thanks! The information is quite useful, and I'll try the functions in Skyline-daily. : )