e juveniles, male broodstock and female broodstock. Brain and hy pophysis from broodstock animals were also dissected and rapidly flash frozen in liquid nitrogen. Gonads were fully isolated in adult and juvenile fish and thus gonadal tissue was devoid of any other tissue. However, gonads of sexually differentiating fish contained Batimastat a bit of attached epithelium. Due to their extremely small size, the isola tion of the gonads alone was not feasible and thus sam ples contained also portions of the surrounding tissues. RNA was individually extracted by RNeasy Mini Kit following the manufac turers instructions. Quantity was determined using a Nanodrop spectrophotometer. The RNA integrity number was deter mined in an Agilent BioAnalizer. RNA samples with a RIN 8. 1 were further processed for the sequencing run.

A pooled sample was generated by mixing 70% of gonads containing equal amounts of RNA from each individual and 30% of equal amount of RNA from broodstock brains and hypophysis tissues. cDNA library, normalization and 454 FLX Titanium pyrosequencing Full length enriched double stranded cDNA was synthe sized from 1. 5 ug of pooled total RNA using the MINT cDNA synthesis kit according to the manufacturers protocol, and was subsequently purified using the QIAquick PCR Purification Kit. The amplified cDNA was normalized using the Trimmer kit to minimize differences in representation of transcripts. The single stranded cDNA fraction was then amplified twice by sequential PCR reactions according to manufacturers protocol. Normalized cDNA was purified using the QIAquick PCR Purification Kit.

Normalized cDNA was used to gen erate a 454 library. cDNA was fractionated into small, 300 to 800 bp fragments and the specific A and B adaptors were ligated to both the 30 and 50 ends of the fragments and used for purification, amplification, and sequencing steps. Two and a quarter PTP regions were used for the GS FLX sequencing run using Titanium chemistry. All reagents and protocols were from Roche 454 Life Sciences, USA. 454 data was processed with Roches software, using default settings, to obtain fasta and quality files containing the trimmed sequence of all reads. Contigs with at least 100 bp were recovered. Sequences were de novo assembled into contigs by running Mira v3. 2. 0rc1 in EST mode. Contigs less than 100 bp were filtered out and the rest was blasted against D.

rerio RefSeq protein sequences with est2assemblys analyse assembly. pl script in order to validate the whole process. Turbot databases Bioinformatic tools were developed to process all sequen cing data obtained from both Sanger and 454 FLX Titanium technologies. The starting point of the current work was the Turbot 1 database, which was reported previ ously. In order to generate the Turbot 2 database se quences of Turbot 1 database were clustered with, 3,043 sequences obtained from the E. scophthalmi trial cDNA libraries, 1,371 genomic sequences from enriched DNA libraries and 3,339 sequences a