I am trying to analyze volatiles for a batch of samples that were acquired using a GC single quadrupole Agilent instrument. We are trying to get minor differences between the samples, so the parameters set had a very low threshold (eg: peak height >10 (slightly above background)). I've run the data through the Profinder software, using both the batch Molecular Feature Extraction and the batch recursive Feature Extraction. I am getting similar results (number of peaks identified) using both wizards, although it seems like the batch MFE is detecting peaks in a smaller number of samples than the batch recursive FE. My questions are:
1) For data obtained by a GC single quad, is one wizard better than the other? I understand that the recommendation is to set the parameters using the batch MFE first before running the batch recursive MF. Is the Find by Ion algorithm useful for deconvoluting peaks with nominal mass? I tried a number of different parameters in the Find by Ion algorithm but it didn't seem to change the number of compound groups identified. Is there any recommended settings?
2) I did go through the list of compounds identified and some of them are definitely assigned wrong. Eg, peaks with a completely different mass spectrum that are 0.5 min apart (I minimized this by setting a smaller RT window). Should I delete this at this step or can I feed this analyzed data through Mass Profiler Professional and correct for the data here?
3) How do I evaluate if Profinder is deconvoluting the peaks correctly?
I appreciate any recommendations!