BRAKER2: automatic eukaryotic genome annotation with GeneMark-EP+ and AUGUSTUS supported by a protein database

Abstract The task of eukaryotic genome annotation remains challenging. Only a few genomes could serve as standards achieved through tremendous investment human curation efforts. Still, the correctness all alternative isoforms, even in best-annotated genomes, be good subject for further investigation. new BRAKER2 pipeline generates and integrates external protein support into iterative process training gene prediction by GeneMark-EP+ AUGUSTUS. continues line started BRAKER1 where self-training GeneMark-ET AUGUSTUS made predictions supported transcriptomic data. Among challenges addressed was generation reliable hints to protein-coding exon boundaries from likely homologous but evolutionarily distant proteins. In comparison with other pipelines annotation, is fully automatic. It favorably compared under equal conditions pipelines, e.g. MAKER2, terms accuracy performance. Development should facilitate solving harmonization genes different species. However, we understand that several more innovations are needed proteomic technologies well algorithmic development reach goal highly accurate genomes.