GFF3toEMBL: Preparing annotated assemblies for submission to EMBL

Abstract

An essential part of open reproducible research in genomics is the deposition of annotated de novo assembled genomes in public archives such as EMBL/GenBank (Blaxter et al. 2016). The interfaces provided by the major archives do not allow for data to be easily submitted on a large scale without substantial prior knowledge on the part of the sub-mitter. This has lead to a situation where less than 15% of all sequenced bacteria have corresponding public assemblies. We address this by providing GFF3toEMBL, which con-verts the output of the most commonly used automatic annotation tool, Prokka (Seemann 2014), and converts it to a format suitable for submission to EMBL. Built on the Genome-Tools annotation processing library (Gremme, Steinbiss, and Kurtz 2013), GFF3toEMBL is robust, fast, memory efficient and well tested, and has been used to submit more than 30% of all annotated genomes in EMBL/GenBank (Page et al. 2016). It is a small, but essential missing step in making genomic research more open and reproducible.