Skip to main content
. 2022 Feb 7;13:785196. doi: 10.3389/fpls.2022.785196

TABLE 8.

Prediction accuracy (ρ) ranges of breeding values for selection of host plant resistance to late blight, tuber yield, starch percentage, and crisp quality in potato using different training population sizes and varying number of testing environments.

Characteristic Training population size (N) and testing environments Prediction method ρ References
Host plant resistance to late blight N = 273; early and advanced breeding clones with records from 7 years at single site using non-replicated plots or three 4-hill plots in randomized complete block design (RCBD) Bayesian ridge regression (BRR), Bayes B 0.24–0.31 Enciso-Rodriguez et al., 2018
N = 184; breeding clones and parents with data scoring over 3 years at single site GBLUP, Bayes A, Bayes Cπ, Bayesian LASSO (BL) 0.32–0.86 Stich and Van Inghelandt, 2018
N = 241–336 at two sites using RCBD GBLUP 0.52–0.68 Gemenet et al., 2020
N = 92; first generation (T1) plus 4 ancestors with data scoring from 1 year at single site BRR, Bayes A, Bayes B, Bayes C, BL 0.13–0.24 Selga et al., 2021a
N = 301; two first generation (T1) half-sib offspring (n1 = 151, n2 = 149) with data from 1 year at single site BRR 0.16–0.31 Selga et al., 2021b
Total tuber weight N = 190; EU released cultivars evaluated in single site over 2 years using RCBD with three replications BL, RKHS, Bayes A, Bayes B, Bayes C ca. 0.25–ca. 0.34 Habyarimana et al., 2017
N = 184; breeding clones and parents with data recording over 3 years at single site GBLUP, Bayes A, Bayes Cπ, BL 0.43–0.55 Stich and Van Inghelandt, 2018
N = 571; data from breeding clones over 6 years GBLUP 0.06–0.31 Endelman et al., 2018
N = 413; T3 offspring in replicated trials along with standard cultivars HBLUP (pedigree, phenotypic, and genomic information) 0.32–0.34 Sood et al., 2020
N = 241–336 at two sites using augmented designs GBLUP 0.16–0.38 Gemenet et al., 2020
N = 665, non-replicated trials of T1 (n1 = 465 in 4-plant plots) and T2 (n2 = 138 in 10-plant plots) at one site plus T3+ (n1 = 62 in 20-plant plots) offspring in replicated trials across three sites BRR 0.05–0.75 Selga et al., 2021b
N = 147; cultivars and very advanced breeding clones with testing across three sites over 2 years GBLUP, BL, Bayes A, Bayes Cπ 0.55–0.59 Wilson et al., 2021
Tuber starch or specific gravity N = 762; hybrid offspring derived from biparental crossing of 18 plus unrelated 74 breeding clones for model validation GBLUP, Bayes A, Bayes C 0.09–0.81 Sverrisdóttir et al., 2017
N = 190; EU released cultivars evaluated in single site over 2 years using RCBD with three replications BL, RKHS, Bayes A, Bayes B, Bayes C ca. 0.13–ca. 0.69 Habyarimana et al., 2017
N = 1,146; mapping population (n1 = 762) over 2 years at one site (non-replicated in year 1 and RCBD with two reps in year 2) plus two testing panels (n2 = 92 incl. 18 parents of mapping population; n3 = 292 breeding clones) with trial data over years GBLUP 0.37–0.71 (across pops) 0.75–0.83 (cross validating) Sverrisdóttir et al., 2018
N = 571; data from breeding clones over 6 years GBLUP 0.13–0.63 Endelman et al., 2018
N = 184; breeding clones and parents with data recording over 3 years at single site GBLUP, Bayes A, Bayes Cπ, Bayesian Lasso 0.51–0.83 Stich and Van Inghelandt, 2018
N = 200; non-replicated T2 (n2 = 138 in 10-plant plots) at one site plus T3+ (n1 = 62 in 20-plant plots) offspring in replicated trials across three sites BRR 0.43–0.62 Selga et al., 2021b
N = 147; cultivars and very advanced breeding clones with testing across three sites over 2 years GBLUP, BL, Bayes A, Bayes Cπ 0.72–0.76 Wilson et al., 2021
Crisp quality (reducing sugars or fry color) N = 762; hybrid offspring derived from biparental crossing of 18 plus unrelated 74 breeding clones for model validation GBLUP, Bayes A, Bayes C 0.16–0.56 Sverrisdóttir et al., 2017
N = 571; data from breeding clones over 6 years GBLUP 0.40–ca. 0.45 Endelman et al., 2018
N = 555 breeding clones evaluated in 20-plant plots in a single year rrBLUP, Bayesian A, Bayesian Lasso, Random Forest 0.11–0.77 (of the field)
0.24–0.66 (after low temperature storage)
Byrne et al., 2020
N = 1,146; mapping population (n1 = 762) over two years at one site (non-replicated in year 1 and RCBD with two reps in year 2) plus two testing panels (n2 = 92 incl. 18 parents of mapping population; n3 = 292 breeding clones) with trial data over years GBLUP 0.28–0.48 (across pops) 0.39–0.79 (cross validating) Sverrisdóttir et al., 2018