Re: [Psidev-ms-dev] Standardize representation for MS^e and AIF

[Chambers, M. <mat...@gm...>]

Hi Eric, you were in the discussion of this issue in the private thread with the Waters folks. Do you have a preference for one of the 
options below, or another to suggest?

On 8/14/2018 5:07 PM, Chambers, Matthew wrote:
> Hi all,
>
> Over the last few months there have been quite a few users asking me about how to distinguish the low and high energy levels in MS^e [1]. 
> It seems most users expect them to correspond to MS1 and MS2 (most of the users were actually reporting all MS^e scans being the same MS 
> level as a bug). As far as I know, reporting both low and high energy MS^e spectra as MS1 is the current standard approach. But it seems 
> obvious that users need a way to distinguish between the low and high energy spectra, so something needs to change with mzML. I see a few 
> options and there may be others.
>
> 1. Keep the existing schema: continue treating low energy spectra as MS1, but treat high energy spectra as MSn (ms2) with an 
> isolationWindow the same width as the scanWindow, and add a bogus selectedIon in the middle of the window.
>
> 2. Change the schema: make isolationWindow and selectedIonList optional so collision energy can be stored in the activation element 
> without specifying bogus window/ion information.
>
> 3. Change the schema: allow the activation element in either the <scan> element or the <precursor> element? This would avoid the need to 
> have a "precursor" for a spectrum which really doesn't have one.
>
> With either option 2 or 3, there are a few aspects to decide:
>   a. Should low energy MS^e spectra be considered MS1 or MSn (ms2)? If MS1, should they still have an activation element in order to 
> preserve collision energy info?
>   b. If low energy spectra are treated as MS1, should high energy spectra be considered MS1 or MSn (ms2)? High energy spectra certainly 
> need the activation element.
>   c. Should MS^e spectra get their own spectrum type instead of trying to shoehorn them into MS1/MSn? Perhaps a general term like 
> "broadband fragmentation spectrum"?
>
> It's not clear to me which option is the best, but if I had to choose I'd probably pick 1. Because it won't force schema-validating 
> parsers to update, and old software will still be able to read the new files (but that would only be a good thing in SOME cases; in other 
> cases the software should probably give an error that it can't handle that kind of data).
>
> This has been an outstanding issue for mzML for a while. I hope we can finally get it resolved.
>
> Thanks,
> -Matt
>
>
> [1] Briefly, MS^e is a data-independent-acquisition method where all ions are fragmented simultaneously rather than isolating any specific 
> precursors. It's up to the analyzing software to make sense of the resulting complex spectra by looking at how the peaks change over time. 
> The technique may or may not be combined with ion mobility separation (HDMS^e ). The spectra are acquired in an every-other-scan scheme, 
> i.e. low-high-low-high-low-high. Low energy spectra have very little collision energy, but it's not exactly zero.