MH Qual - Find by molecular feature

Question

We would like to establish a workflow for the semi-quantification of unkowns in our samples. To do so, we measure a calibration row of a bunch of external standards and our sample. Internal standard is added by co-injection directly in the multisampler. 
 After being measured, we use MH Qual to identify (unknown) compounds in our samples. To do so, we use the FBMF workflow. When we tested the workflow with our lowest calibration standard (10 ppb), we encountered a problem: Even though we tried to adjust the quality score and all other parameters in several different combinations, our calibration substance is not recognized as a feature. In contrast to that, when we extract an EIC, we see a nice peak of the substance. 
 Now my question: Given the low LOQ, which we cannot increase, is it somehow possible to improve the FBMF workflow to find this feature? Or how else would you tackle this? 
 We are also reluctant to use the FBF workflow, since we want to do a non-target screening. 
 After the Qual workflow, we would the import the data (via .cef) into Quant to do a semi-quantification. 
 We use an Agilent QTOF 6546 Software: MassHunter Qualitative Analysis Version 10.0 Service Release 1

howard_sanford · Answer

Hello bofzgi84 , 
 Find by Molecular Feature works in the following way (taken from the Qual Help) 
 This algorithm treats all of the mass spectral data from the experiment as a large, three-dimensional array of (retention time, m/z, abundance) values. It removes from that array any points that correspond to persistent or slowly-changing background. Then it searches for features that have a common elution profile (for example, for masses that elute at very nearly the same time). Those masses are then further grouped into one or more "compounds" containing m/z values that are related (for example, they correspond to peaks in the same isotope cluster, or can be explained as being different adducts or charge states of the same entity). 
 One possibility then is that while you can extract a clean EIC of the molecular ion, if the corresponding isotope clusters or other signals that might be used to group it are not strong, then it may be missed by the algorithm. 
 I would start with a higher concentration sample where the compound can be found and compare the signals observed there to the concentrations where the compound is not found. For adjusting parameters, I would adjust settings on the tabs starting with the left most bottom row tab and working to the right and then the back tabs left to right. This is generally the order that all of the method settings are applied during the workflow. 
 
 You could try adjusting the settings in the Compound ion count threshold section to allow for fewer and lower abundance ions. Though when experimenting you may need to restrict the RT range to just the compound you are trying to adjust to avoid unnecessary processing time for the complete run.

howard_sanford · Answer

Hello bofzgi84 , 
 As an update to this, a colleague of mine has pointed out that for LC data the Compound ion count threshold settings are not present. I took the above screenshot with no data file loaded, so all possible options were enabled. I can get that dialog to appear for LC data if I override the data type and specify GC and LC for the interface configuration, but for LC data it ignores the setting and appears to default to only needing 2 ions to find a feature. 
 It looks like you can force it to accept compounds with only one ion by enabling the option in the Charge state section to report unassignable ions as having a charge state of 1. 
 
 I did test this with the LCMS demo data file Test_PosMS.d. Without this option 56 compounds were found with default settings. With this option enabled an additional 254 single ion compounds were found. Adjusting the setting to something a little more reasonable for this data, a 7,500 count peak filter and a 1% relative height Compound filter, I get 14 compounds total, 7 of which are single ion. 
 
 Since this could increase the number of compounds found and the time it would take to process a data file, you may want to turn off the option to Extract complete result set automatically on the Results tab. This will still find and show all compounds, but it will not extract the chromatograms and spectra for a compound until you click on it in the list. This option also controls how data files are loaded when the Load result data option is enabled in the file open dialog, so it can speed the loading of processed data as well.