The effect of marker types and density on genomic prediction and GWAS of key performance traits in tetraploid potato

Abstract

Genomic prediction and genome-wide association studies are becoming widely employed in potato key performance trait QTL identifications and to support potato breeding using genomic selection. Elite cultivars are tetraploid and highly heterozygous but also share many common ancestors and generation-spanning inbreeding events, resulting from the clonal propagation of potatoes through seed potatoes. Consequentially, many SNP markers are not in a 1:1 relationship with a single allele variant but shared over several alleles that might exert varying effects on a given trait. The impact of such redundant "diluted" predictors on the statistical models underpinning genome-wide association studies (GWAS) and genomic prediction has scarcely been evaluated despite the potential impact on model accuracy and performance. We evaluated the impact of marker location, marker type, and marker density on the genomic prediction and GWAS of five key performance traits in tetraploid potato (chipping quality, dry matter content, length/width ratio, senescence, and yield). A 762-offspring panel of a diallel cross of 18 elite cultivars was genotyped by sequencing, and markers were annotated according to a reference genome. Genomic prediction models (GBLUP) were trained on four marker subsets [non-synonymous (29,553 SNPs), synonymous (31,229), non-coding (32,388), and a combination], and robustness to marker reduction was investigated. Single-marker regression GWAS was performed for each trait and marker subset. The best cross-validated prediction correlation coefficients of 0.54, 0.75, 0.49, 0.35, and 0.28 were obtained for chipping quality, dry matter content, length/width ratio, senescence, and yield, respectively. The trait prediction abilities were similar across all marker types, with only non-synonymous variants improving yield predictive ability by 16%. Marker reduction response did not depend on marker type but rather on trait. Traits with high predictive abilities, e.g., dry matter content, reached a plateau using fewer markers than traits with intermediate-low correlations, such as yield. The predictions were unbiased across all traits, marker types, and all marker densities >100 SNPs. Our results suggest that using non-synonymous variants does not enhance the performance of genomic prediction of most traits. The major known QTLs were identified by GWAS and were reproducible across exonic and whole-genome variant sets for dry matter content, length/width ratio, and senescence. In contrast, minor QTL detection was marker type dependent.

Keywords: GBLUP; GWAS; Solanum tuberosum; genomic prediction; marker density; marker type; tetraploid potato breeding.

Figure 1

Heat map of marker density in 1-Mb windows for each chromosome. (A) Non-synonymous variants, (B) non-coding variants, (C) synonymous variants, and (D) combination set of annotated variants. The color gradient denotes marker count.

Figure 2

Distribution of phenotypes in the MASPOT panel. (A) Chipping quality, (B) length/width ratio, (C) senescence, (D) dry matter content, and (E) yield.

Figure 3

Principal component analysis of the genomic relationship matrix of the annotated combination set of 93,170 SNP markers for 755 MASPOT panel clones colored by (A) mother and (B) father. Principal component 1 is plotted against principal component 2 in the top and against principal component 3 in the bottom. The principal components explain 19.62%, 10.8%, and 9.5% of the total explained variance, respectively.

Figure 4

GBLUP prediction correlation coefficient and bias within the MASPOT panel over a number of annotated markers. The markers were reduced from full annotated SNP data sets by iterative reduction to every other marker (every 10th for the final reduction) to avoid the introduction of positional bias. Each reduction was repeated up to four times for each iteration. Color by marker annotation (SNPtype). (A) Prediction correlation coefficient between observed and predicted phenotypic values. Boxplots of correlation coefficients determined for each marker count with connecting lines through the mean. (B) Bias of GEBVs estimated as the slope of the linear regression line between observed and predicted phenotypic values. Boxplots of biases determined for each marker count with connecting lines through the mean bias.

Figure 5

Q–Q plots of expected -log₁₀(p-values), assuming no marker–trait association, against observed (chipping quality and the full combination set in this example). (A) Before correction with the genomic inflation factor and (B) after correction.

Figure 6

Manhattan plots of -log₁₀(p-value) GWAS results of the full 93,170 marker annotated combination set. (A) Chipping quality, (B) length/width ratio, (C) senescence, (D) dry matter content, and (E) yield. Chromosome 0 is pseudomolecules. The significance threshold with Bonferroni correction is indicated with a horizontal line.