This would involve developing some sense of peak shape in the IM dimension, but we could probably start with something as simple as the comparing (IM at the maximum intensity point across the IM filter band at the peak in the RT dimension) vs (nominal IM of the ion).

John F at Agilent feels this is a highly useful piece of information.

Per JohnF, comparing against the IM value at the intensity center of gravity across the purple band is the way to go. Expressed as % error in IM and % error in CCS.

Probably you would want to be consistent with what we do in the m/z dimension, which is to use the weighted mean of the signal in the extraction range. This often ends up being just a single value in centroided data, but it also works well in messier profile mode data, like profile-mode TOF m/z. I think it likely works better when resolution decreases and an apex value is not clearly the center of the distribution. This method has the ability to give a value that is between two buckets with relatively similar signal.

I think this has been requested a number of times since it was originally posted. Most recently, this support request on PanoramaWeb.org suggests it would be useful to track in AutoQC.

https://panoramaweb.org/home/support/announcements-thread.view?rowId=9801