High-quality SNPs from genic regions highlight introgression patterns among European white oaks (Quercus petraea and Q. robur).

In the post-genomics era, non-model species like most Fagaceae still lack operational diversity resources for population genomics studies. We completed the analyses of Sanger sequences produced from over 800 gene fragments covering ~530 kb across the genic partition of European oaks in a range-wide sampling of 25 individuals (11 Quercus petraea, 13 Q. robur, one Q. ilex as an outgroup). Regions targeted represented broad functional categories potentially involved in species ecological preferences, and a random set of genes. Using a high-quality dedicated pipeline, we provide a detailed characterization of over 14500 polymorphisms, including ~12500 SNPs -218 being triallelic-, over 1500 insertion-deletions, and ~200 novel di- and tri-nucleotide SSR loci. This catalog also includes various summary statistics within and among species, gene ontology information, and standard formats to assist loci choice for genotyping projects. The distribution of nucleotide diversity and differentiation across genic regions are also described for the first time in those species (mean ΘΠ close to ~0.0049 in Q. petraea and to ~0.0045 in Q. robur across random regions, and mean FST ~0.13 across SNPs). Robust patterns were observed which emphasize a slightly but significantly higher diversity in Q. petraea, across a random gene set and in the abiotic stress functional category, and a heterogeneous landscape of both diversity and differentiation. These patterns are discussed in the context of both species documented introgression history despite strong reproductive barriers. The quality, representativity in terms of species genomic diversity, and usefulness of the resources provided are discussed for possible applications in medium scale landscape ecology projects, and as a reference resource for validation purposes in larger-scale re-sequencing projects. These are preferentially recommended in oaks in contrast to SNP array development, given the large nucleotide variation and low levels of linkage disequilibrium revealed.