In my PhD-work, we are trying to develop LC-MS/MS methods for targeted analysis. Usually, I determine my quantifying ion/fragment based on a few different criteria:

  • If there is a 13C internal standard of identical structure, stick with the same fragment for both.
  • If not, look for the lowest noise/highest area/lowest LOD
  • If two fragments are similar in the previous manner, I have a look if there are different levels of contamination in the method blanks

For most compounds, that works, but for a small subset I have the case where the method blanks are similar, but for the fragment with the lowest LOD, noise is highest and/or areas are lowest between the two fragments we analyze for.

Would you go for the fragment with the highest area, highest noise (assuming a large number is more robust against small changes) and the highest LOD, or the lowest one? Is there a best practice I can refer to? There is a 1-2 order of magnitude difference between the LODs, so I really don't want to misrepresent the data.

Thanks in advance!


1 Answer 1


The limit of detection (LOD) in these kinds of assays depends on both the noise and the signal. Usually it is defined as lowest concentration analyzed such that the signal (peak height) is three standard deviations higher than the noise. Exactly how to define "the noise" is a critical consideration -- many different sources provide different definitions and ideas. The best measurement of "noise" is probably in a method blank at the exact RT window you are using for quantification. Ideally the method blank would have a sample matrix identical to your real samples, but guaranteed to be missing the target compound.

It's a little tough to know from the information provided in your question how exactly you are calculating your LODs. If for the moment we assume you are happy with how your LODs are computed, then I would pick the method with the lowest LOD.

It's also worth mentioning that a standard practice in targeted LC-MS/MS is to monitor two separate transitions per compound. One of those is called the "quantifier" and it is this signal that is used to compute peak areas for quantification. The other is called the "qualifier". The ratio of the qualifier peak area to the quantifier peak area should be monitored for all analyses, and it should be approximately the same for all samples (that contain measuremable compound). If this ratio deviates wildly, it's a sign that the quantifier ion may not be reliable for a particular sample.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.