Selection of trait-specific markers and multi-environment models improve genomic predictive ability in rice

Abstract

Developing high yielding rice varieties that are tolerant to drought stress is crucial for the sustainable livelihood of rice farmers in rainfed rice cropping ecosystems. Genomic selection (GS) promises to be an effective breeding option for these complex traits. We evaluated the effectiveness of two rather new options in the implementation of GS: trait and environment-specific marker selection and the use of multi-environment prediction models. A reference population of 280 rainfed lowland accessions endowed with 215k SNP markers data was phenotyped under a favorable and two managed drought environments. Trait-specific SNP subsets (28k) were selected for each trait under each environment, using results of GWAS performed with the complete genotype dataset. Performances of single-environment and multi-environment genomic prediction models were compared using kernel regression based methods (GBLUP and RKHS) under two cross validation scenarios: availability (CV2) or not (CV1) of phenotypic data for the validation set, in one of the environments. Trait-specific marker selection strategy achieved predictive ability (PA) of genomic prediction up to 22% higher than markers selected on the bases of neutral linkage disequilibrium (LD). Tolerance to drought stress was up to 32% better predicted by multi-environment models (especially RKHS based models) under CV2 strategy. Under the less favorable CV1 strategy, the multi-environment models achieved similar PA than the single-environment predictions. We also showed that reasonable PA could be obtained with as few as 3,000 SNP markers, even in a population of low LD extent, provided marker selection is based on pairwise LD. The implications of these findings for breeding for drought tolerance are discussed. The most resource sparing option would be accurate phenotyping of the reference population in a favorable environment and under a managed drought, while the candidate population would be phenotyped only under one of those environments.

Fig 1. Boxplot of phenotypic variables within the rice population.

E1c and E1: data from non-stressed lowland environment with 280 and 204 accessions, respectively. E2 and E3: data from lowland-drought and upland-drought environments, respectively, with 204 accessions.

Fig 2. Predictive ability of genomic prediction for combinations of five levels of linkage disequilibrium (LD) and three levels of minor allele frequency (MAF) in cross validation experiments in the rice population of 280 accessions, under non-stress condition (E1), for days to flowering (DTF), grain yield (GY) and plant height (PH), obtained with 2 statistical methods, GBLUP and RKHS.

Fig 3. Predictive ability of genomic prediction in cross validation experiments implemented with 28,091 SNP derived using two marker selection methods: Linkage disequilibrium (LD) between markers, white boxes.

Genome wide association analysis with the target traits (GWAS), green boxes. The three traits, days to flowering (DTF), grain yield (GY) and plant height (PH), were phenotyped under three environments: rainfed lowland (E1), rainfed lowland with drought stress (E2) and upland with drought stress (E3). For each box, the mean (x) and median (horizontal bar) values are represented.

Fig 4. Predictive ability of genomic prediction experiment with single environment (SE), and multi-environment (ME) models obtained with the GBLUP, RKHS-1 and RKHS-2 statistical methods.

Traits studied are days to flowering (DTF), grain yield (GY) and plant height (PH). The ME models are implemented with two cross-validation strategies CV1 and CV2. Three environments are considered: lowland no-drought (E1, blue), lowland drought (E2, gray) and upland drought (E3, orange). Environment in brackets contributed to the training of ME models. For each box, the mean (x) and median (horizontal bar) values are represented.