Asymmetric peaks: They look pretty, but they don't have a bearing on how efficient a quantitation is.
Most people misunderstand peak area. It is essentially the sum of peak height slices (which are electrical current measurements) and there are various ways to cut the Gaussian pie.
Therefore, if chromatography is consistent, asymmetric peak or not, then there should be no effect on quantitation.
In fact, with mass spectrometry, the slower your scan, depending on peak width, the less likely you will have a symmetric peak. A poorly sampled symmetric peak can appear asymmetric. In that case, your peak asymmetry factor would convey a false notion.
So, let's say the peak has a tail: That would be 5-10% error at best, which is within instrument error on the calibration curve (usually < 30% RSD tolerance).
If an analyte is quantified via peak area ratios, specifically in isotopic dilution methods (see image LEFT), any distortion of native analyte peak shape should be mirrored by an identical distortion in its isotopically-labelled analogue and the ratio of these two peaks will most likely still fall within the ratio control limits (depending on how distorted the peak is). The peaks in the image to the left are in control according to QC (take my word for it)!
Chromatographic separation requirements are more important than peak symmetry requirements.
Detection limit requirements are more important than peak symmetry requirements.
Quality control requirements are more important than peak symmetry requirements.
As long as the data are in control, that is more important than peak symmetry.
If you are "removing" acceptable data because it's asymmetric, then you are removing data that is in control simply because it doesn't "look good". And that doesn't look good!
"Thank you Pat," I mutter. "This makes my job a lot easier!"
~ Filtered by Dave Hope ~
