Skip to main content
Topic: Verifying MS1 Identification Accuracy (Read 336 times) previous topic - next topic

Verifying MS1 Identification Accuracy

Hello all,

I am working on lc-ms/ms data using MS-DIAL v.4.9.

My samples are animal tissues but I found some plant metabolites in MS1 after identification.

How do we check whether the identification of these MS1 metabolites was accurate or not?

Thank you.

 

Re: Verifying MS1 Identification Accuracy

Reply #1
Hi KCNP,

If you only have MS1 information (accurate mass in a scan and retention time), you will have difficulties evaluating your annotations further, without more analysis.

Based on your described information (MS1, RT, Name or Structure) alone;
- First, you can use the predicted elemental composition based on accurate mass and isotope pattern. It might disagree with your annotation by the library (your plant metabolite). But to be honest, that doesn't get you very far and as far I know is implemented into mzMine already.
- Second, there are some tools/ways to estimate retention times, which might help you exclude your annotation.
You can also make a crude estimate yourself using logP values and molecular surface (e.g. lipophilic cholesterol shouldn't elute early in reverse-phase chromatography on a C18 column). However, you will need to understand your method and elution behaviours well before making any such conclusions and they usually are crude estimates. 
- You can also search for isomers of your plant metabolite. Maybe there is a common animal metabolite known to be an isomer to your plant metabolite.

So, what else can you do?
  - One option is a re-injection of your sample (or a pooled QC) which shows your plant metabolite, and use a targeted MS2 acquisition list to obtain fragmentation spectra (MS2). Then you can try to match/dismatch it to spectral libraries.
  - With an MS2 in hand you might even attempt some structure elucidation (SIRUS, etc.)
  - If MS2 is not an option (you might run on a low-resolution single quad MS), you could get a reference standard for those compounds. However, that seems like a lot of (financial) effort to evaluate possible annotations at this level. If that's not an issue, because you already have a standard or it's cheap, run it and compare your retention time information.

As a further note, you might find plant metabolites in some animal tissue, e.g. through diet or because they are animal and plant metabolites.
And remember to clearly communicate your annotation level for all metabolites, even if they fit nicely into your study (animal metabolites). Just because something fits into your hypothesis/tissue/study doesn't mean its also correctly annotated, especially on MS1.

Good luck with your study and welcome to the frustrating but fascinating world of metabolomics annotation. ;)