breed ing strategies. The advents of new high throughput se quencing technologies, which produce extensive sequence data, are providing new opportunities to increase the amount of molecular markers, as demonstrated in the stur geon, where hundreds of SNPs were discovered. Overall, the improvement of the turbot aquaculture in dustry by selecting, on one hand, the most resistant broodstock and, on the other hand, female biased batches is a priority challenge. The purpose of this study was to in crease turbot database information for genes related to the immune and reproductive systems by creating a powerful tool for genomic research in this species. The turbot data base was updated with genes obtained both by Sanger se quencing from immune related tissues after challenges with the myxozoan parasite E.
scophthalmi and by a 454 FLX Titanium run from gonad and brain hypophysis at different stages of development. Description and compari son of the two sequencing strategies, annotation proce dures, Drug_discovery and construction of a larger database, the support for microsatellites and SNP discovery, and for designing a pilot microarray platform, are presented. Results and discussion The increase of known immune related genes in turbot by Sanger sequencing The progression in the construction of the turbot data base is summarized in Table 1. First, the Turbot 1 data base was created from almost ten thousand high quality EST sequences from three cDNA libraries of three im mune relevant organs generated from turbot infected with A. salmonicida sub species salmonicida and P.
dicentrarchi, as well as from non infected fish. The Turbot 2 database included several resource sequences, i 1,371 sequences from seven microsatellite enriched DNA libraries from muscle tissues, ii 3,339 ESTs available in public databases, which were loaded on the turbot database and clus tered with the set of the existing EST, and iii Sanger se quencing data from two new cDNA libraries generated from several immune tissues after challenging with the myxosporean parasite E. scophthalmi produced a total of 3,043 sequences. Together, Sanger based sequencing generated 17,626 sequences with an average length of 501 base pair, constituting the Turbot 2 database. The assembly of all these available data consisted of 6,170 putative transcripts of which 1,827 were contigs and 4,343 singletons.
A high level of redundancy was found, which is usually observed when non normalized cDNA libraries are used, but it constitutes an appropriate approach to obtain a first picture of the im mune response. A total of 6,053 out of the 6,170 unique sequences in Turbot 2 database displayed significant matches with sequences available in public databases with E values equal or less than 1,00E 5. Gene Ontology annotation classified sequences as follows, 586 in Biological Process, 472 in Cellular Component and 692 in Molecular Func tions. 454 pyrosequencing of the turbot brain hypophysis gonad axis transcriptome The Turbot 2