Re: [Psidev-pi-dev] Issue 42 in psi-pi: Issues with the CV

[Martin E. <mar...@ru...>]

> Do we allow internal nodes of the PSI:PI ontology show up in an axml
> instance?
> 
> E.g. at
> /AnalysisXML/AnalysisProtocolCollection/ProteinDetectionProtocol/AnalysisParams
> the
> mapping file currently specifies that only the child terms of
> "PI:00194 - quality estimation with decoy database" are allowed. However,
> the MPC
> example uses exactly this term, to say that a quality estimation was done
> using a
> decoy database. What I like to do is moving all the children of PI:00194 up
> into "search input details -> quality estimation options" and keep the term as an
> indicator that quality estimation was done by decoy database.
At this position in the CV I can imagine to allow only child terms.
The MPC example is redundant as
it uses "PI:00194 - quality estimation with decoy database" and
"PI:00197 - forward+reverse decoy DB" and
"PI:00283 - decoy DB accession regexp".

The first (non-child) term could be dropped, so no changes are necessary.

What we need for the example is another term "decoy DB generation method" or
"decoy DB generation algorithm". 
I will propose that separately in issue 42.
 
  Bye
    Martin