The output of this pipeline is a set of genome-specific hints and, optionally, an in-silico gene prediction. However, in most cases this is only the first step in an annotation project. Here are some suggestions on where to go next.
Basic gene structure and synteny is often conserved across closely related organisms. This information can thus be used to further improve the prediction of gene models.
Using AUGUSTUS, it is possible to run such a comparative gene build using hints from the related genomes of interest and a whole-genome alignment graph to identify syntenic regions for gene prediction. For more information, please see here.
A simple functional annotation can be generated by "blasting" the predicted protein sequences (part of the output of this pipeline) against a Uniprot/Swissprot formatted reference database. A tool that makes this workflow quite easy is Annie.
Automated gene prediction will typically contain numerous errors that are best rectified through manual review and curation. A suitable platform for this is available with WebApollo.