Expanding the BLUP alphabet for genomic prediction adaptable to the genetic architectures of complex traits

Jiabo Wang,Zhiwu Zhang,Edward S Buckler,Peter J Bradbury,Hui Li,Zhengkui Zhou,Zhe Zhang,Di Liu,Qin Zhang

doi:10.1038/s41437-018-0075-0

Abstract

Improvement of statistical methods is crucial for realizing the potential of increasingly dense genetic markers. Bayesian methods treat all markers as random effects, exhibit an advantage on dense markers, and offer the flexibility of using different priors. In contrast, genomic best linear unbiased prediction (gBLUP) is superior in computing speed, but only superior in prediction accuracy for extremely complex traits. Currently, the existing variety in the BLUP method is insufficient for adapting to new sequencing technologies and traits with different genetic architectures. In this study, we found two ways to change the kinship derivation in the BLUP method that improve prediction accuracy while maintaining the computational advantage. First, using the settlement under progressively exclusive relationship (SUPER) algorithm, we substituted all available markers with estimated quantitative trait nucleotides (QTNs) to derive kinship. Second, we compressed individuals into groups based on kinship, and then used the groups as random effects instead of individuals. The two methods were named as SUPER BLUP (sBLUP) and compressed BLUP (cBLUP). Analyses on both simulated and real data demonstrated that these two methods offer flexibility for evaluating a variety of traits, covering a broadened realm of genetic architectures. For traits controlled by small numbers of genes, sBLUP outperforms Bayesian LASSO (least absolute shrinkage and selection operator). For traits with low heritability, cBLUP outperforms both gBLUP and Bayesian LASSO methods. We implemented these new BLUP alphabet series methods in an R package, Genome Association and Prediction Integrated Tool (GAPIT), available at http://zzlab.net/GAPIT.

Highlights

These authors contributed : Jiabo Wang, Zhengkui Zhou and Zhe ZhangElectronic supplementary material The online version of this article contains supplementary material, which is available to authorized users.One of the ultimate goals of genomic research is to predict phenotypes from genotypes
Kinship based on true or estimated quantitative trait nucleotides (QTNs). To show how these two new best linear unbiased prediction (BLUP) variations work and their advantages over both genomic best linear unbiased prediction (gBLUP) and Bayesian LASSO, we conducted a series of studies on both simulated phenotypes and real phenotypes in three species (Arabidopsis, mice, and maize)
To fully understand which BLUP or Bayes method should be used with different genetic architectures, we examined prediction accuracies on simulated traits

Summary

1234567890();,: 1234567890();,: Introduction

Improvement on gBLUP’s prediction accuracy was found using kinship derived from weighted markers This method was named trait-specific relationship matrix (TA) BLUP (taBLUP) (Zhang et al 2011). Where y is a vector of phenotypes; β represents unknown fixed effects, including population structure and associated Quantitative Trait Loci (QTLs); and u is a vector of genomic prediction with size n (the number of individuals) for unknown random polygenic effects These random effects follow a distribution with a mean of zero and a covariance matrix of G 1⁄4 Aσ2a1⁄42Kσ2a, where K = 0.5 A is the pedigree-based kinship with element Kij (i, j = 1, 2, ..., n) representing the relationship between individuals i and j, A = 2 K is additive numerator relationship, and σ. The predicted phenotypes of inference were compared to their true phenotypes to evaluate prediction accuracy

Results

Discussion

CONCLUSION

Compliance with ethical standards