W328
A New Look at Population Recombination and Linkage Disequilibrium in Forest Trees from Genome-Wide SNP Data in Eucalyptus and its Relevance to Molecular Breeding

Date: Sunday, January 10, 2016
Time: 11:10 AM
Room: Sunrise - Meeting House
Orzenil Bonfim Silva-Junior , Bioinformatics Lab - EMBRAPA Genetic Resources & Biotechnology, Brasilia, Brazil
Dario Grattapaglia , Plant Genetics Lab - EMBRAPA Genetic Resources & Biotechnology, Brasília, DF, Brazil
Understanding the genome-wide patterns of recombination provides important insights into the shaping of the species' genetic history and on the extent of linkage disequilibrium (LD), which in turn determines our ability to dissect quantitative traits and predict complex phenotypes. We characterized the population-scaled recombination rate (ρ) and LD in E. grandis using 21,517 polymorphic genome-wide Infinium SNPs and nearly 13 million SNPs from whole-genome pooled resequencing of 36 individuals. At the genome-wide scale, LD decayed within ≈4-6 kb, considerably slower than previously reported from candidate gene studies, but showing variation from absence to complete LD up to 50-kb. A sharp decrease in the estimate of ρ was seen when going from short (0-100 bp) to genome-wide (0-50 kb) inter-SNP distances, highlighting the dependence of ρ on the scale of observation, while further corroborating that LD estimates from candidate gene studies cannot provide accurate expectations of genome-wide LD. Our results challenge the heretofore-established consensus of rapid decay of LD in outcrossed forest tree genomes, converging with other recent genome-wide studies in Poplar. The more extensive LD found has considerable practical importance to molecular breeding. While a genome-wide usable LD within ≈4-6 kb in natural populations warrants a positive outlook on the prospects of GWAS from the detection standpoint, it complicates pinpointing causative QTNs, if such a feat will ever be accomplished. On the other hand, with extensive LD in closed breeding populations, low-density genotyping platforms will provide abundant power to capture linked effects for whole-genome prediction at tree breeders’ accessible costs.