Résumé
New generation sequencer (NGS) allows to obtain within only one run hundreds of billions bases in order to study SNP diversity. The first step of this work we develop lab protocols adapted for NGS to test them for grape. In this context we have firstly developed a nuclear DNA extraction protocol to limit the level of cytoplasmic DNA, and secondly with have analyzed three runs of 454 GS-FLX. Our protocol allows the limitation of cytoplasmic DNA at BAC libraries similar level (4%) whereas a classical DNA extraction contains around 12%. Our data sets from the 454 have been aligned on the grape reference genome PN40024 (12X) with a compilation of different software: Mosaik-assembler, Repeat-Masker. Sixty-six percent of the reads obtained from each 454 runs aligned at a unique locus (30% coverage of the grape genome of reference). The quantity of reads aligned is proportional of the length of chromosomes. Compare to the PN40024 genome, we obtained a similar GC percentage, coding and non coding ratio, element mobile ratio. Nerveless, we observed that some genomic regions are more easily accessed by these types of methodologies. The chromosome structure, centromeres and histones localizations could be responsible for this observation. Such protocol would be a great asset for sequencing project as GrapeReseq.