Cassava yield traits predicted by genomic selection methods

Abstract

Genomic selection (GS) has been used to optimize genetic gains when phenotypic selection is considered costly and difficult to measure. The objective of this work was to evaluate the efficiency and consistency of GS prediction for cassava yield traits (Manihot esculenta Crantz) using different methods, taking into account the effect of population structure. BLUPs and deregressed BLUPs were obtained for 888 cassava accessions and evaluated for fresh root yield, dry root yield and dry matter content in roots in 21 trials conducted from 2011 to 2016. The deregressed BLUPs obtained for the accessions from a 48K single nucleotide polymorphism dataset were used for genomic predictions based on the BayesB, BLASSO, RR-BLUP, G-BLUP and RKHS methods. The accessions’ BLUPs were used in the validation step using four cross-validation strategies, taking into account population structure and different GS methods. Similar estimates of predictive ability and bias were identified for the different genomic selection methods in the first cross-validation strategy. Lower predictive ability was observed for fresh root yield (0.4569 –RR-BLUP to 0.4756—RKHS) and dry root yield (0.4689 –G-BLUP to 0.4818—RKHS) in comparison with dry matter content (0.5655 –BLASSO to 0.5670 –RKHS). However, the RKHS method exhibited higher efficiency and consistency in most of the validation scenarios in terms of prediction ability for fresh root yield and dry root yield. The correlations of the genomic estimated breeding values between the genomic selection methods were quite high (0.99–1.00), resulting in high coincidence of clone selection regardless of the genomic selection method. The deviance analyses within and between the validation clusters formed by the discriminant analysis of principal components were significant for all traits. Therefore, this study indicated that i) the prediction of dry matter content was more accurate compared to that of yield traits, possibly as a result of the smaller influence of non-additive genetic effects; ii) the RKHS method resulted in high and stable prediction ability in most of the validation scenarios; and iii) some kinship between the validation and training populations is desirable in order for genomic selection to succeed due to the significant effect of population structure on genomic selection predictions.

Factors Influencing Genomic Prediction Accuracies of Tropical Maize Resistance to Fall Armyworm and Weevils

“…This supposed genetic architecture difference between FAW and MW-resistance trait could be the reason for non-linear methods such as random forest performing better at predicting FAW resistance, since these are more capable of integrating epistasis in the statistical modelling [ 27 , 51 ]. However, the RKHS algorithm, also a non-linear GP approach known to efficiently handle epistatic genetic relation [ 51 , 59 ], did not successfully run on FAW dataset, although it was among the best models for predicting MW-resistance traits, except BLUPs for the number of affected kernels (AK), for which the RKHS algorithm did not run successfully. In this study, the reasons for some GP algorithms failing to run either on MW or FAW-resistance datasets are unclear, but this could be related to the BLUPS structure of the datasets that failed to run.…”

“…Compared to the above-mentioned parametric methods, deep learning techniques such as support vector regression (SVR), multilayer perceptron, and convolutional neural networks models performed poorly in some studies [ 55 , 56 ]. However, there are also instances where RKHS outperformed one or several of the parametric methods, for instance, GBLUP, rrBLUP, and Bayesian algorithms, in terms of several traits in several crops including maize [ 51 , 57 , 58 , 59 ]. These results were most likely because nonparametric GS models capture more adequately the non-additive genetic components which are an essential characteristic of complex traits [ 23 , 37 , 38 ] and hence could be good candidate tools for the prediction of FAW and MW-resistance traits which are controlled by both additive and non-additive gene action [ 21 , 23 , 31 , 41 , 60 ].…”

Selecting for useful properties of plants and fungi – Novel approaches, opportunities, and challenges

“…soya, maize, wheat and cotton) to increase yields and disease resistance, and hence improve crop quality (Crossa et al., 2017; González‐Camacho et al., 2018; Michel et al., 2018; Rutkoski et al., 2015), while proof of concept has also been demonstrated for quality traits, such as the baking qualities of bread wheat (Michel et al., 2018). In the staple food crop cassava, in which phenotypic selection alone is inefficient and heritability of desired traits is low, genomic selection has been shown to accurately predict yield traits in multiple trials (de Andrade, Sousa, Oliveira, Resende, & Azevedo, 2019). Nevertheless, the presence of polyploidy and high heterozygosity in many crop genomes is currently limiting the widespread application of genomic selection (Friedmann et al., 2018).…”

BRS Novo Horizonte - a new cassava variety for industrial use

“…to Andrade et al 2019); and dry root yield (t ha -1 , calculated as the dry matter content multiplied by the fresh root yield). The harvest index (in %, obtained by the ratio between fresh root yield and total biomass-shoot and root weight) was analyzed as a secondary trait in the selection process.…”