Annotation for ballan wrasse (Labrus bergylta)
A two-pass iteration with MAKER version 2.31.8 was performed on the final genome assembly as described at https://github.com/sujaikumar/assemblage/blob/master/README-annotation.md and in Campell et al with some minor modifications. First, three ab initio gene finders were trained, SNAP version 20131129on the genes found by CEGMA version 2.4.010312, GeneMark-ES version 2.3e on the genome assembly itself, and AUGUSTUS version 3.0.2 trained with the Actinopterygii gene set from BUSCO v1.1b by using the –long option for BUSCO. SwissProt/UniProtKB (release 2016_01), and used as input to MAKER together with two sets of transcriptomes (one assembled with Trinity, and one based on 454 reads), the repeat library and the trained gene finders for the first pass of training. SNAP and AUGUSTUS were then retrained on the GFF output of MAKER, and a second pass was performed. Gene ontologies and gene families and domains were applied to the protein output of the second pass by running InterProScan version 5.4-47, and putative gene names were identified by BLAST against SwissProt/UniProtKB. Gene models with AED of more than 0.5 were removed, resulting in a final set of gene models numbering 31,050.
The files with aed_0.5 in their names, are the filtered. The others are the whole annotation.
