Search

Cassava (Manihot esculenta L.) is one of the most important food crops in the tropics, but yields are far below their potential. The gene pool of cas- sava contains natural genetic diversity relevant to many important breeding goals, but breeding progress has been slow, partly because of insuf- ficient genomic resources. As a first step toward implementing genomewide genetic studies that will facilitate rapid genetic gain through breed- ing, we genotyped-by-sequencing a set of 182 full-sibs population of cassava that segregated in several traits: resistance to the cassava mosaic disease (CMD) and yield under CMD pressure; increased carotenoid content in storage roots; color of stem exterior and anthocyanin pigmen- tation in the petioles, inner root skin, and api- cal leaves. Employing a rare-cutting restriction enzyme, PstI, in a genotyping-by-sequencing (GBS) library preparation, we obtained 2478 segregating single nucleotide polymorphisms (SNPs), of which 1257 passed standard filter - ing for missing genotypes and deviation from expected genotypic frequencies. We mapped 772 SNPs across 19 linkage groups and anchored 313 unique scaffolds from the version 4.1 of the cassava genome assembly. Most of the stud- ied morphological traits as well as resistance to CMD and root carotenoid content showed quali- tative inheritance. As expected, quantitative trait loci analysis for these traits revealed single loci surrounded by small confidence intervals. Yield under CMD was associated with the CMD resis- tance locus. We show that GBS is a powerful genotyping tool that provides a sufficient number of markers for unraveling the genetic architecture of Mendelian traits in cassava in addition to the development of a robust genetic map that can help anchor unassembled genomic scaffolds.

Before implementation of genomic selection, evaluation of the potential accuracy of prediction can be obtained by cross-validation. In this procedure, a population with both phenotypes and genotypes is split into training and validation sets. The prediction model is fi tted using the training set, and its accuracy is calculated on the validation set. The degree of genetic relatedness between the training and validation sets may infl uence the expected accuracy as may the genotype × environment (G×E) interaction in those sets. We developed a method to assess these effects and tested it in cassava (Manihot esculenta Crantz). We used historical phenotypic data available from the International Institute of Tropical Agriculture Genetic Gain trial and performed genotyping by sequencing for these clones. We tested cross-validation sampling schemes preventing the training and validation sets from sharing (i) genetically close clones or (ii) similar evaluation locations. For 19 traits, plot-basis heritabilities ranged from 0.04 to 0.66. The correlation between predicted and observed phenotypes ranged from 0.15 to 0.47. Across traits, predicting for less related clones decreased accuracy from 0 to 0.07, a small but consistent effect. For 17 traits, predicting for different locations decreased accuracy between 0.01 and 0.18. Genomic selection has potential to accelerate gains in cassava and the existing training population should give a reasonable estimate of future prediction accuracies.