The Effects of Demography and Long-Term Selection on the Accuracy of Genomic Prediction with Sequence Data

Abstract

The use of dense SNPs to predict the genetic value of an individual for a complex trait is often referred to as "genomic selection" in livestock and crops, but is also relevant to human genetics to predict, for example, complex genetic disease risk. The accuracy of prediction depends on the strength of linkage disequilibrium (LD) between SNPs and causal mutations. If sequence data were used instead of dense SNPs, accuracy should increase because causal mutations are present, but demographic history and longterm negative selection also influence accuracy. We therefore evaluated genomic prediction, using simulated sequence in two contrasting populations: one reducing from an ancestrally large effective population size (N e ) to a small one, with high LD common in domestic livestock, while the second had a large constant-sized N e with low LD similar to that in some human or outbred plant populations. There were two scenarios in each population; causal variants were either neutral or under long-term negative selection. For large N e , sequence data led to a 22% increase in accuracy relative to 600K SNP chip data with a Bayesian analysis and a more modest advantage with a BLUP analysis. This advantage increased when causal variants were influenced by negative selection, and accuracy persisted when 10 generations separated reference and validation populations. However, in the reducing N e population, there was little advantage for sequence even with negative selection. This study demonstrates the joint influence of demography and selection on accuracy of prediction and improves our understanding of how best to exploit sequence for genomic prediction. M ETHODOLOGY has been developed to predict genetic value for polygenic traits in livestock and crops by exploiting high-density genome-wide SNP genotypes that are fitted simultaneously in an analytical model (Meuwissen et al. 2001). The same methodology can be applied in human genetics, for example, to predict complex disease risk (reviewed in De los Campos et al. 2010). In livestock and plants this analytical approach is often referred to as "genomic selection" because the genomic predictions are used for selection decisions. First, a large "reference population" with genotypes and phenotypes is required to jointly estimate genome-wide SNP effects. Then the accuracy of prediction using the estimated SNP effects is reevaluated in an independent "validation population," before the genomic prediction equation is routinely applied on individuals with genotypes but no phenotypes.Genomic prediction (GP) methods generally use dense genome-wide SNP genotypes and therefore rely on exploiting linkage disequilibrium (LD) between these SNPs and unknown causative mutations or quantitative trait loci (QTL). The lower the LD is between the SNP and causal mutations, the lower the accuracy will be of GP. As the number of generations separating the reference and validation populations increases, the LD between SNPs and causative mutations is further eroded by recombination and t...

Toward genomic prediction from whole-genome sequence data: impact of sequencing design on genotype imputation and accuracy of predictions

“…To inform the simulation of sequence data for the founders of this population, we relied on the work of Macleod et al (2012A), which used multilocus patterns of LD in whole-genome sequence from two Holstein founder bulls (Larkin et al, 2012) to reconstruct population genetic history. Based on this population history (with variable effective population size), Macleod et al (2012B) simulated sequence data that closely matched the multilocus LD patterns in the founder bull sequences. We observed that patterns of multilocus LD in dense SNP data from Belgian Blue cattle were very similar to Holsteins.…”

“…We observed that patterns of multilocus LD in dense SNP data from Belgian Blue cattle were very similar to Holsteins. Therefore, we used data simulated by Macleod et al (2012B) with Fregene (Chadeau-Hyam et al, 2008) as sequence data for founders of our Belgian Blue cattle population. Figure 1 describes the overall simulation scheme.…”

Review: Opportunities and challenges for small populations of dairy cattle in the era of genomics

“…However, as has been shown by BrØndum et al (2015), a change in accuracy may be limited to a 1% to 5% increase. MacLeod et al (2014) showed that a benefit from the use of sequence data mainly can be achieved in populations with a large effective population size and/or a comparatively low level of LD while in populations like the Holstein population, an increase of accuracy for genomic breeding values may only be very small. Both conditions, a large effective size and a low level of LD, hence are not fulfilled for typical small populations as seen in livestock.…”

Evaluating the accuracy of genomic prediction of growth and wood traits in twoEucalyptusspecies and their F1hybrids

“…A number of simulation and empirical studies have shown that increasing the number of markers may improve the predictive accuracy as N e also increased [9,[21][22][23]. However, increasing the number of markers in small N e populations provides little or no improvement on predictive accuracy [24,25].…”