[Psidev-ms-dev] Centroid vs. Profile for mass spectra

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

Hi all,

I'm trying to map mzXML's various ways of indicating profile vs. 
centroid mode spectra to mzML. The mapping would work best if the 
optional dataProcessing attributes ("centroided" and "deisotoped", which 
refer to the entire file) in mzXML map to the fileContent element in 
mzML. But MS_centroid_mass_spectrum and MS_profile_mass_spectrum are not 
currently "data file content" types so they would not validate if they 
were in fileContent.

It looks like the semantic validator requires that either centroid or 
profile mass spectrum appear in each spectrum_description (if I'm 
reading the ms-mapping file correctly), so that's good. The only thing a 
converter from an ambiguous-data-mode file (like DTA/MGF/possibly mzXML) 
to mzML can do is assume that the data is one way or the other. I 
suppose it could do some heuristics to guess intelligently, too.

-Matt