Linkage Disequilibrium Analysis in a Two-Generation Pedigree of Maritime Pine

Linkage disequilibrium (LD) is one of the main factors affecting the accuracy of genomic predictions. Indeed, genomic prediction relies on large number of DNA markers that cover the whole genome to exploit the linkage disequilibrium between markers and any quantitative trait loci. To study the extent of LD in Pinus pinaster L., a two-generation population (184 unrelated founders selected from natural plantations and 477 descendants) was genotyped using 2,500 SNPs. Among them, 2,184 were assigned at a genetic map location on the 12 linkage groups (LG) of the maritime pine composite map. Intra-chromosomal LD was estimated using both r² and r²_v which is corrected from relatedness among genotypes using realized genomic relationship matrix. Average LD (r²) for all LGs ranged from 0.008 to 0.019 and were significantly smaller than 0.1 (p-value << 0.01). For unbiased LD (r²_v), the range was from 0.005 to 0.012. The major differences between the two estimates of LD occurred for closely linked marker (< 20 cM). After correction for genetic relatedness in the population, the LD (r²_v), 876 pairs of markers had an LD greater than 0.6. About 91% involved physically linked SNPs belonging to the same contig, while 5.4% involved completely genetically linked SNPs (0 cM). The remaining 3.6% were related to bias in composite linkage map. The overall mean LD for the genome was 0.011 using r² and it was 0.006 using r²_v suggesting, a rapid decay in a short distance as expected for an outbred species with large effective population size.