Utilizing evolutionary conservation to detect deleterious mutations and improve genomic prediction in cassava

Abstract

Introduction: Cassava (Manihot esculenta) is an annual root crop which provides the major source of calories for over half a billion people around the world. Since its domestication ~10,000 years ago, cassava has been largely clonally propagated through stem cuttings. Minimal sexual recombination has led to an accumulation of deleterious mutations made evident by heavy inbreeding depression.

Methods: To locate and characterize these deleterious mutations, and to measure selection pressure across the cassava genome, we aligned 52 related Euphorbiaceae and other related species representing millions of years of evolution. With single base-pair resolution of genetic conservation, we used protein structure models, amino acid impact, and evolutionary conservation across the Euphorbiaceae to estimate evolutionary constraint. With known deleterious mutations, we aimed to improve genomic evaluations of plant performance through genomic prediction. We first tested this hypothesis through simulation utilizing multi-kernel GBLUP to predict simulated phenotypes across separate populations of cassava.

Results: Simulations showed a sizable increase of prediction accuracy when incorporating functional variants in the model when the trait was determined by<100 quantitative trait loci (QTL). Utilizing deleterious mutations and functional weights informed through evolutionary conservation, we saw improvements in genomic prediction accuracy that were dependent on trait and prediction.

Conclusion: We showed the potential for using evolutionary information to track functional variation across the genome, in order to improve whole genome trait prediction. We anticipate that continued work to improve genotype accuracy and deleterious mutation assessment will lead to improved genomic assessments of cassava clones.

Keywords: cassava (Manihot esculenta); deleterious mutation; evolutionary conservation; genetic load; genomic prediction.

Figure 1

Species Alignment Depth Across Cassava Genes. Alignment depth represented by the number of species with homologous alleles in each multiple sequence alignment at any given protein coding base pair in the cassava genome.

Figure 2

Defining Deleterious Mutations. (A) baseml evolutionary rate is plotted against SIFT scores. Deleterious mutations were classified as derived alleles at those sites with a baseml evolutionary rate < 0.5 and a SIFT score < 0.05 (Black box). (B) Distribution of homozygous and heterozygous deleterious mutations across 1048 cassava clones.

Figure 3

Predicted Functional Weights. Histogram of functional weights produced through RandomForest prediction of conservation for nonsynonymous variant sites. High functional weights correspond to highly conserved sites where nonsynonymous mutations are predicted to have large functional effects.

Figure 4

Simulated QTL Effects. Histograms show count of QTL effects in one example simulation. Each facet shows a genetic architecture with different proportions of the markers acting as QTL (resulting in ~ 6600, 660, 66, and 6 QTL on average). The x-axis represents the positive effect of carrying the ancestral allele at a given QTL.

Figure 5

Genomic Prediction Accuracies with Simulated QTL. Prediction accuracies are shown on the y-axis as the correlation between predicted andtrue breeding values. The x-axis delineates the prediction scenario being tested. Barplot color corresponds to the genomic information used in the prediction model. Error bars represent a 95% confidence interval for simulations. Simulations were repeated with different proportions of the markers acting as causative QTL: 0.1 (A), 0.01 (B), 0.001 (C), and 0.0001 (D).

Figure 6

Fresh Root Yield Genomic Prediction Leveraging Deleterious Annotations. Prediction accuracy is measured in cross-population and within-population prediction scenarios. Genomic models are represented as bar graph colors where various genomic and deleterious data are used in the genomic prediction. Error bars represent a 95% confidence interval for within-population 10-fold prediction.

Figure 7

Dry Matter Percentage Genomic Prediction Leveraging Deleterious Annotations. Prediction accuracy is measured in cross-population and within-population prediction scenarios. Genomic models are represented as bar graph colors where various genomic and deleterious data are used in the genomic prediction. Error bars represent a 95% confidence interval for within-population 10-fold prediction.