Skip to main content
Topic: Some GC-MS peaks donot correctly aligned (Read 338 times) previous topic - next topic

Some GC-MS peaks donot correctly aligned


I found that some GC-MS peaks do not align correctly. For example, benzaldehyde and 1.2.4-trimethylbenzene have quite difference spectra according to the NIST library, however they are aligned as one peak and only 1.2.4-trimethylbenzene is reported even though I have set 80% spectrum similarity for the alignment (please see the attachment). Does anybody know what is wrong here?

thanks in advance.

Re: Some GC-MS peaks donot correctly aligned

Reply #1
Hi Sukis,

Maybe, these metabolites have similar RTs, right?
Maybe, the result depends on the reference file that you selected.
If the reference file contains EI-MS spectra mixing these two metabolites, the two metabolites in each file can be aligned into one aligned spot.
Did you try to do the following things?
1. in the MSP library, set different QUANTMASS values for each metabolite
2. then, process the files with the MSP file.

Please let me know your update.


Re: Some GC-MS peaks donot correctly aligned

Reply #2
Dear Hiroshi,

I am awfully sorry for the late reply. I just forgot about this issue.....
It is true that these two compounds have quite close RI. I just thought that MSDIAL will take into account the spectra similarity as well.
setting quant mass for each spectrum and use quant masses defined in MSP format file do the trick. So maybe it is better to set quant mass for each spectra in the *.msp file? but if i understand well, MSDIAL will take the highest peak in each spectra as the quantmass, right?