You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have a new assembly for a genome, and I have developed a new annotation (using a combination of Mikado - Augustus - PASA). I also did a liftover of the old genome assembly annotation. Looking at them in the browser, there's some improved models in the new annotation, and some better solved in the older version.
I think being able to run Mikado with two annotations, including all the scoring information from portcullis introns, blast results etc, would be a good way of picking the best gene models for every locus, instead of simply going with the annotation with highest BUSCO score. But so far I don't think that's possible with the current Mikado configuration (it only takes one reference annotation). Since the filters and scoring are pretty much all there, do you think it would be possible to add this function?
Thanks in advance for considering this suggestion.
Best,
Alex de Mendoza
The text was updated successfully, but these errors were encountered:
Following this post, I'm also interested in using Mikado to update a set of 'ab initio' generated reference transcripts. Currently, we use Mikado to generate the hints that we use with Augustus to predict genes. However, we lose UTRs and isoforms with Augustus. I would thus like to use Mikado to update the Augustus transcripts with the RNA-seq data (stringtie gtfs and Trinity mapped transcripts with gmap). I see there are options to do so, like using --only-reference-update at the pick stage, but the outputs I'm getting so far are quite suboptimal, and I'm unsure if I'm doing things right. I would greatly appreciate any guidance!
Dear Mikado developers,
I have a new assembly for a genome, and I have developed a new annotation (using a combination of Mikado - Augustus - PASA). I also did a liftover of the old genome assembly annotation. Looking at them in the browser, there's some improved models in the new annotation, and some better solved in the older version.
I think being able to run Mikado with two annotations, including all the scoring information from portcullis introns, blast results etc, would be a good way of picking the best gene models for every locus, instead of simply going with the annotation with highest BUSCO score. But so far I don't think that's possible with the current Mikado configuration (it only takes one reference annotation). Since the filters and scoring are pretty much all there, do you think it would be possible to add this function?
Thanks in advance for considering this suggestion.
Best,
Alex de Mendoza
The text was updated successfully, but these errors were encountered: