Re: [Psidev-pi-dev] [Psidev-ms-dev] storing intermediate results in mzML

[Matthew C. <mat...@va...>]

Richard Scheltema wrote:
> Dear all,
>
> Thank you for your comments. A question from my side, is the phone 
> conference scheduled for this afternoon open? As I'm on the agenda I 
> would like to join if possible.
>   
Yes, the call is open to all interested parties. BYOB. :)


> Matthew Chambers wrote:
>   
>> I'm not sure what this would look like. We certainly didn't have 3d 
>> chromatograms in mind, but perhaps they can be accommodated. Would that 
>> be a chromatogram with three axes (data arrays): time, m/z, and 
>> intensity? Is this akin to the "psuedo-2d-gel" view?
>>     
> The mass chromatogram would indeed consist of time, m/z and intensity. 
> The m/z data coming from the profile spectrum data. If you plot the mass 
> chromatogram it will look a bit like a bell. I'm not quite sure what you 
> mean with the pseudo-2d-gel view, but as I understand it, this is a 2d 
> view of all the mass chromatograms found in for example a LC/MS profile. 
> In the definition of the chromatogram block some semantics should be 
> worked out in order to let the parser figure out what is stored there.
>
> I would suggest something as following (a mass chromatogram spanning 9 
> scans, sorry for the verbosity of it all). For the centroid data the 
> original structure would hold with the addition of the cvParam MS:1000127:
>
> <snip>
I fear this is a pretty cryptic representation that would rarely be 
supported. I'm thinking the preferable way to store these chromatograms 
is simply as a series of 2d chromatograms. Yes, it means repeating the 
metadata and the time array every time, but I don't feel that's a 
critical issue. Actually, repeating the metadata is probably a bonus, 
because each slice of the 3d chromatogram (the separated dimension being 
m/z) should correspond with a single m/z. We have yet to settle on the 
"right" way to put m/z in the chromatogram elements, even for simpler 
concepts like an SRM chromatogram.


>> Yes, I think a background ion chromatogram is reasonable. But there may 
>> be some semantics to work out: is there a distinction between 
>> "background ions" and "noise"? I know there are different kinds of noise 
>> to think about...
>>     
> There is a distinction between noise and background ions. The definition 
> for a background ion should according to me in LC/MS be a compound which 
> is continuously eluting from the LC/GC or continuously being injected 
> into the MS. A famous example of this would be PEGs dissolving from the 
> plastic tubing used to transport the sample from the LC to the MS. So 
> background ions are real (usable) compounds, while noise are artefacts 
> from the measurement technique used not describing a real compound. 
>   
I would argue that both electronic noise and chemical ("real") noise are 
kinds of noise. They are also both ways to get a certain level of 
background intensity. I glean these meanings from the "signal to noise" 
calculation often used. As far as I know, the "noise" denominator there 
refers to the sum of electronic and chemical noise. So in my lexicon, 
noise and background are nearly synonymous, but "electronic noise" and 
"chemical noise" are distinct causes of noise. A "background ion 
chromatogram" would also, I expect, refer to the sum of electronic and 
chemical noise. How would you separate them?


> Isomers forming a peak on top of the elution pattern of a background ion 
> should probably be removed.
>   
I'm not sure I understand this statement; does it refer to some data 
processing on the raw chromatogram to remove signal when it is 
co-eluting with noise?


> If this definition is maintained I think the addition to the CV that I 
> proposed still holds, but I could be wrong and more is necessary? 
> Probably single spectra (no separation up front) would suffer a bit, 
> although then we're not really talking about a chromatogram right?
>
>  name: backgroundion chromatogram
>  is_a: chromatogram type (MS:1000626)
>  definition: chromatogram created by creating an array of a ubiquitously present mass.
>   
I'm used to chromatograms referring to either a specific m/z (selected 
ion chromatogram), the total ion current, or some mass range within each 
spectrum. In the last case, the intensity within the mass range is 
summed to create a single value. Does this definition invoke the 
"specific m/z" usage by choosing an m/z known to be a source of chemical 
noise/background intensity? And if so, we're talking about a specific 
kind of selected ion chromatogram?

-Matt