Description
VC2010 gff is an outlier in terms of it's content as it is a "special" tier III genome. It contains mappings of N2 annotations onto the VC2010 genome. The drop is mainly down to the script being reworked and where you don't get a gene:cds:protein:peptide we now don't print out the objects. For all non-vc2010 species this essentially has no change as the gff and fasta files are in sync, for VC2010 this causes a dramatic reduction in the number of objects created. Following investigation this is ok. However since we no longer add the bioproject to the gene/object stubs the N2 genes in the VC2010 gff get printed out with an updated Public_name which tramples over the canonical N2 WBGene objects.
Cleaning the VC2010 stubs more, and see if the current gene set is sensible, or if the script autoace_builder.pl -postmerge = tier3_stubs.pl needs a few more subroutines to handle the VC2010 gene names