Helixer: trained animal and plant models for cross species gene annotation

Trained models associated with the paper "Helixer: Cross-species Gene Annotation Of Large Eukaryotic Genomes Using Deep Learning" (under review), and the code base: https://github.com/weberlab-hhu/Helixer. All models for the final ensemble are included here.

In short: these models can be used to make base-pair wise gene annotation predictions (whether each base in the genome belongs to an intergenic, untranslated, coding, or intronic region). Validation efforts indicated that the animal models perform well for cross-species predictions within vertebrates, while the plant models perform well within Embryophyta.

model_info.csv contains some brief meta information and model performance on the validation set.

The best single animal model (if running one model and not an ensemble) is animals_a_e07.h5, while the best single plant model is plants_a_e10.h5.