4 Replies Latest reply on Nov 19, 2018 6:20 AM by trifecta

    MS MH Quant Sharpness Threshold


      Using MH Quant B.08.  In Library Method>Edit>Deconvolution tab, there is a Component shape - sharpness threshold parameter. 

      What is this supposed to do ?

      I have tried multiple values to try and deconvolute 2 peaks that are partially resolved, with no luck.  I have also tried mutiple different combinations of resolution settings, with no luck.  The peaks successfully deconvolute when I use AMDIS, just with default settings.

        • Re: MS MH Quant Sharpness Threshold



          I added tags to this post to increase visibility.

          • Re: MS MH Quant Sharpness Threshold

            The Resolution (RT window size factor) controls how much "separation" there must be between two partially coeluting compounds to make them a single compound or separate compounds.  This is somewhat like the Resolution setting in AMDIS.  If you make the number larger it will merge compounds with close RTs, if you make it smaller it will separate them.  If you make it too small it will split a single compound apart based on very slight RT differences in the individual ions (EIC peaks).  


            The Sharpness threshold is used to select which ions in a compound will be used to create the model peak shape for the compound.  There are many coeluting ions in the compound that can be selected so the software will pick a percentage of the sharpest peaks.  Sharpness is basically a measure of how wide the peak is compared to how high it is.  Sharper peaks are better as they are less likely to be affected by a coeluting compounds for any given ion.  As a practical matter I would leave the default value in place.


            Sometimes AMDIS can separate a peak when MassHunter's deconvolution does not.  One difference that I have seen is when the two compounds have very similar RTs but different chromatographic peak shapes (Example: tailing in only one compound).  AMDIS considers peak shape while Deconvolution does not.  However without the data this is only a guess.