Exome Genotyping and Association Mapping in Conifers: Focusing on Norway Spruce

The main goal was to implement exome capture in Norway spruce (Picea abies) to increase the number of known single nucleotide polymorphisms (SNPs) and genetic resources available for research and tree breeding. The exons were captured in 521 mother trees using in house designed probes. 80 000 hybridization probes were designed from the Norway spruce genome 1.0. SNP detection was performed using GATK haplotype caller using scaffolds from the 80 000 probe design after initially mapping to the whole Norway spruce genome v1.0. A total of 545 546 high quality SNPs were acquired. After filtering the data using MAF 0.05 and Missingness 20% 160 366 SNPs remained. The SNPs were then used for association mapping studies of wood chemistry traits, which also include growth, density separated into latewood and earlywood. Estimated breeding values from 5618 trees of the 521 open-pollinated families in two 21-year-old Norway spruce progeny trials in southern Sweden were used as the phenotype data for all the 21 traits studied. The growth trajectories of the phenotype data were determined as non-linear. A quadratic spline model was then applied in order to define the latent traits. Association mapping was performed using a multilocus LASSO penalized regression method and a stability selection method in order to calculate the stability selection probability of each SNP being selected to the model. With the first five principle components calculated based on the genotype data being included as covariates into the model to take the population structure into account. The threshold for declaring a SNP to be significant or suggestive was 0.6976 and 0.5395, by maximum allowing 1 and 5 false positive selections, respectively. Association studies has already yielded SNPs significantly associated with density and growth, our focus now turning to using this method to perform association analysis phenology datasets.