Consider extension of metadata to contain analysis-type

[nilshoffmann]

In order to select an assay type in MetaboLights (which differs in naming / semantics) from mzTab-M, we need to indicate via a cv parameter, what type of analysis / technology was used to measure the MS data. E.g. GC-MS, LC-MS, etc.

At the moment, this kind of information is only indirectly available within mzTab-M from the quantification_method or optional CV Terms.

Such an element could be placed at the mzTab-M files metadata top level. However, mzTab-M supports multiple instrument results in one file. Thus, we would need to establish a link to the corresponding instrument sections. In order to keep the format simple,
multiple assay types (in MetaboLights terminology) should be reported as separate mzTab-M files (e.g. a separate one for GC-MS, LC-MS, HILIC, etc.) and the analysis type should be reported in the metadata custom cv parameters list.

We need to amend the specification to clearly indicate that results from multiple assay types (e.g GC-MS, LC-MS, but also RPLC vs Hilic) should be placed in different mzTab-M files. Each mzTab-M file should then report the MTBLS-analysis-type as a hint to the translation engine to map this to an individual MTBLS ISA structure.

To indicate the type of assay to use, the analysis type can be defined as follows (where it occurs is not important, if multiple custom terms are available):

MTD    custom[1]   [,, MTBLS-analysis-type, "LCMS"]

LCMS, GCMS, CEMS, DI (direct infusion), MSI (MS Imaging)

Ideally, these terms will be formalised as CV terms in collaboration with MetaboLights and MetabolomicsWorkbench.