Recently-sequenced non-model organisms often have a set of gene models that evolve quite quickly, and it is easy to lose track of why particular genes were added, removed or modified. Checking final models into NCBI is the endpoint, but it isn't feasible if you are doing repeated rounds of gene model adjustment. Checking our gene models into version control allows us to track who editied which models and why (assuming that people leave sensible commit messages).
This assembly was manually annotated using data from proteomics, transcriptomics and microarray studies. We are currently preparing a manuscript for publication. If you use these annotations in your own work, we kindly ask that you cite