Skip to main content
Topic: Options for Noise Cleaning of MS2 spectra for Build an Inhouse Spectral Library? (Read 2792 times) previous topic - next topic

Options for Noise Cleaning of MS2 spectra for Build an Inhouse Spectral Library?

Hi,

we are going to built an MSMS-library in MS-Dial (MSP-file containing MSMS spectra of reference compounds analyzed on our MS device).

Is there a possibilty to do a noise cleanup of acquired MS2 spectra during data processing? Here, I only found one parameter "Relative Abundance Cut-Off" (Identification Reiter). From my point of view this parameter only determines the way the spectral matching is done but this is not a cleaning of MS2 spectra itself. So, the only way would be to curate each single MS2 spectrum manually?

Thanks for your support.

Regards Falco

Re: Options for Noise Cleaning of MS2 spectra for Build an Inhouse Spectral Library?

Reply #1
Using CorrDec will help. I recommend reading the tutorial for creating an AMRT library. This was very helpful to me. CorrDec requires multiple injections at different intensities to deconvolute spectra. The tutorial says to do a 7 point serial dilution with 1 compound/vial, but this was too costly for us, so instead we did 10 non-isobaric compounds per vial and did injection series with increasing injection volume.

Good luck!

-Triston

Re: Options for Noise Cleaning of MS2 spectra for Build an Inhouse Spectral Library?

Reply #2
Hi Triston,

thanks for sharing the workflow.
You mentioned "deconvolution of spectra". I suppose you analyzed your samples by DIA, right? We are using instead DDA. We are choosing one vial per compound. For each vial we have five injections by applying five different collision energies.

Is the workflow you mentioned only applicable for DDA, too?

When I look at the intensities of my fragments MS2 spectra I regularily observe the same low value for several fragments, i.e. pure noise. From my opinion, it would be nice, to have the ability to set a cut-off with respect to absolute intensity in MS2 Dial during data processing.

Best regards
FalcoB