no seq_region, so exclude from hack in generic_fetch().

  If called for a Bio::EnsEMBL::DBSQL::BaseFeatureAdaptor object, add
  the seq_region and coord_system tables to the table list and join the
  main feature table to coord_system.species_id through the seq_region
  table.

Don't delete the GeneGC entry from attrib_type, since if it is deleted it will be re-created later in the script with a different attrib_type_id, which will be inconsistent with that in the attrib_type.txt file.

Added name and GO term projections for 6 new species - dolphin, hyrax, tariser, megabat, kangaroo rat, alpaca.

Add BINARY flag to meta.meta_value to make it case sensitive when doing UNIQUE comparisions (especially for Gallus/gallus)

species_id column in the meta table.

Unit-tests passes.

database in new().

Swap accession and label data in xref table and leave linkage_xref column blank for direct_xref table

Allow score column in protein_feature to be null. Needed for user upload and to be consistent with protein_align_feature and dna_align_feature.

translation id can be passed when a new ProteinFeature is created so it can be later stored in the database (protein_feature table)