Issue 812: Calculate and store peak skewness/asymmetry value

Status:open
Assigned To:Nick Shulman
Type:Todo
Area:Skyline
Priority:3
Milestone:21.2
Opened:2021-05-28 by Josh Eckels
Changed:2021-06-14 by Brendan MacLean
Resolved:
Resolution:
Closed:
2021-05-28 Josh Eckels
Title»Calculate and store peak skewness/asymmetry value
Assigned To»Brendan MacLean
Type»Defect
Area»Skyline
Priority»3
Milestone»21.2
Following up on our email thread, we've had a request recently from Thermo to calculate additional values for each chromatogram peak to assess is asymmetry or skewness. At a high level, the intent is to assess how the observed peak differs from a Gaussian distribution. Once Skyline is calculating the values, we'll want to update Panorama to import it and likely make it available as a QC metric.

I am following up to get more details on reference calculations or other guidance. But if nothing else this issue can serve as a reminder to follow up again.

Here's a related request from 2019:

https://skyline.ms/announcements/home/support/thread.view?rowId=41007

Also note that this is one of the attributes that SProCoP says it tracks:

https://skyline.ms/skyts/home/software/Skyline/tools/details.view?name=SProCoP

2021-05-28 Josh Eckels
TypeDefect»Todo

2021-06-02 Josh Eckels
Kurtosis is also of interest. I'm out of my depth here, but both skewness and kurtosis have multiple variants, some of which are simpler to calculate.

https://en.wikipedia.org/wiki/Skewness#Other_measures_of_skewness

https://en.wikipedia.org/wiki/Kurtosis#Other_measures

apQuant is an example of an application that provides these calculations:

https://ms.imp.ac.at/index.php?action=apQuant

2021-06-14 Brendan MacLean
Assigned ToBrendan MacLean»Nick Shulman
Let's try to have a serious look at what peak statistics we can still add and when. I feel like there is a backlog that we have been hesitant to include because they take up so much space in our files. Maybe we could expose a check-list of extra peak statistics we could calculate and store to allow users to decide whether they want to use the space or not, or something that would keep us from exploding our size for large DIA runs, but still allow more targeted experiments to have a richer set of statistics, as they are requesting.