My lab recently picked up Helixer and have found that it works extremely well. So far the only thing it has performed poorly on is Gymnosperm genomes. Admittedly, testing has been somewhat limited here, I've ran a publicly available Gnetum genome, a not yet public Ephedra genome, and am in the process of testing on Giant Redwood.
Assembly
Busco C Ref
Busco C Helixer
Gnetum
90.6%
76.3%
Ephedra
82.8%
48.0%
Redwood
TODO
TODO
...Any thoughts/advice? What data would be needed to fine tune the land plant model for gymnosperms or train a new model entirely?
My lab recently picked up Helixer and have found that it works extremely well. So far the only thing it has performed poorly on is Gymnosperm genomes. Admittedly, testing has been somewhat limited here, I've ran a publicly available Gnetum genome, a not yet public Ephedra genome, and am in the process of testing on Giant Redwood.
...Any thoughts/advice? What data would be needed to fine tune the land plant model for gymnosperms or train a new model entirely?