[Biojava-dev] (no subject)

[Keith James]

>>>>> "Matthew" == Matthew Pocock <matthew_pocock@yahoo.co.uk> writes:

[...]

    Matthew> Imho it's such a royal pain that it should be changed pre
    Matthew> 1.3 release. Now, what should we change it to?

Well, SequenceDBInstallation uses database IDs to identify databases,
so I'd go with the suggestion posted earlier in the thread i.e. use
the ID instead.

Then SeqSimilaritySearchResult would go from

public interface SeqSimilaritySearchResult
{
    SymbolList getQuerySequence();
    SequenceDB getSequenceDB();
    Map getSearchParameters();
    List getHits();
}

to

public interface SeqSimilaritySearchResult
{
    SymbolList getQuerySequence();
    String getDatabaseID();
    Map getSearchParameters();
    List getHits();
}

There's a similar situation with the getQuerySequence() method. This
forces the user to make a SequenceDB of query Sequences available to
the builder when creating SeqSimilaritySearchResults. I think this
could go the same way, giving a final interface

public interface SeqSimilaritySearchResult
{
    String getQuerySequenceID();
    String getDatabaseID();
    Map getSearchParameters();
    List getHits();
}

This doesn't define what the query sequence ID means as we don't know
which database (if any) it came from. However, I think it's better
than what we have now in that the programmer has the choice of whether
to resolve the ID(s).

My 0.020601 Euro

Keith

-- 

- Keith James <kdj@sanger.ac.uk> bioinformatics programming support -
- Pathogen Sequencing Unit, The Wellcome Trust Sanger Institute, UK -