Journal Pre-proof

Using genomic data to improve the estimation of general

combining ability based on sparse partial diallel cross designs in
maize

Xin Wang, Zhenliang Zhang, Yang Xu, Pengchen Li, Xuecai

Zhang, Chenwu Xu

PII: S2214-5141(20)30084-2
DOI: https://doi.org/10.1016/j.cj.2020.04.012
Reference: CJ 487

To appear in: The Crop Journal

Received date: 29 December 2019

Revised date: 3 May 2020
Accepted date: 31 May 2020

Please cite this article as: X. Wang, Z. Zhang, Y. Xu, et al., Using genomic data to improve
the estimation of general combining ability based on sparse partial diallel cross designs in
maize, The Crop Journal (2020), https://doi.org/10.1016/j.cj.2020.04.012

This is a PDF file of an article that has undergone enhancements after acceptance, such
as the addition of a cover page and metadata, and formatting for readability, but it is
not yet the definitive version of record. This version will undergo additional copyediting,
typesetting and review before it is published in its final form, but we are providing this
version to give early visibility of the article. Please note that, during the production
process, errors may be discovered which could affect the content, and all legal disclaimers
that apply to the journal pertain.

© 2020 Published by Elsevier.

 Journal Pre-proof

Using genomic data to improve the estimation of general combining ability

based on sparse partial diallel cross designs in maize

Xin Wanga, b, Zhenliang Zhangb, Yang Xub, Pengchen Lib, Xuecai Zhangc, Chenwu Xub,*
a
College of Information Engineering, Yangzhou University, Yangzhou 225009, Jiangsu, China
b
Key Laboratory of Plant Functional Genomics of the Ministry of Education/Jiangsu Key Laboratory of Crop Genomics and
Molecular Breeding/Jiangsu Co-Innovation Center for Modern Production Technology of Grain Crops, College of Agriculture,
Yangzhou University, Yangzhou 225009, Jiangsu, China
c
International Maize and Wheat Improvement Center (CIMMYT), Mexico D.F. 06600, Mexico

Abstract: Evaluation of general combining ability (GCA) is crucial to hybrid breeding in maize. Although the
complete diallel cross design can provide an efficient estimation, sparse partial diallel cross (SPDC) is more

of
flexible in breeding practice. Using real and simulated data sets of partial diallel crosses between 266 maize

inbred lines, this study investigated the performance of SPDC designs for estimating the GCA. With different

ro
distributions of parental lines involved in crossing (called random, balanced and unbalanced samplings), different
-p
numbers of hybrids were sampled as the training sets to estimate the GCA of the 266 inbred lines. In this process,

three statistical approaches were applied. One obtained estimations through the ordinary least square (OLS)
re
method, and the other two utilized genomic prediction (GP) to estimate the GCA. It was found that the coefficient
lP

of determination of each approach was always higher than the heritability of a target trait, showing that the GCA

for maize inbred lines could be accurately predicted with SPDC designs. Both the GP approaches were more
na

accurate than the OLS, particularly in the scenario for a low-heritability trait with a small sample size.

Additionally, prediction results demonstrated that a big sample of hybrids could greatly help improve the accuracy.
ur

The random sampling of parental lines had little influence on the average accuracy. However, the prediction for
Jo

lines that never or seldom involved in crossing might suffer from much lower accuracy.

Keywords: General combining ability; Maize; Genomic prediction; Partial diallel; Accuracy

1. Introduction

The concept of general combining ability (GCA), originally defined by Sprague and Tatum [1], refers to the

average performance of a line in hybrid combinations. It can be estimated using the difference between the

average of its hybrids and the general average for all crossings. GCA is mainly a measure of additive effects,

which can be directly transmitted from parents to offspring. In recent years, doubled haploid (DH) technology has

facilitated the generation of a large number of inbred lines, making GCA evaluation the major bottleneck in hybrid

*
Corresponding author: Chenwu Xu, E-mail address: [email protected].
Received: 2019-10-28; Revised: 2020-03-02; Accepted: 2020-05-31.

1
 Journal Pre-proof
maize breeding [2]. Therefore, the evaluation of GCA is a crucial process for hybrid development, and the maize

line with high GCA is an essential component for producing elite hybrids [3].

Estimation of GCA can be obtained easily using the complete diallel cross design or North Carolina Design II

(NC II) [4, 5]. However, with the development of breeding programs, a great quantity of inbred lines are available.

The number of possible crosses grows very rapidly, making these designs time and resource intensive. Partial

diallel cross designs in which only a subset of possible crosses is performed are more attractive options [6]. They

allow the evaluation of a greater number of inbred lines in crosses [7]. In the classic circulant designs [8], each of

n lines is only crossed with s other lines, instead of (n  1) lines as in the complete diallel. In this way, each line is

guaranteed to be involved in the sampling crosses. Analysis using circulant diallels significantly reduces the

of
number of crosses in which each genitor is involved, and enables the participation of a greater number of genitors

[9]. However, in breeding practice, only a small proportion of hybrids are possible to be evaluated in the field,

ro
making the partial diallel table very sparse. This means that the average value of s is small, and some complex
-p
factors of field trials may bring big fluctuation to s for the parental lines. In this scenario, limited research has

been reported to estimate the GCA. So, there is an urgent need to develop procedures to allow the accurate
re
evaluation of GCA based on sparse partial diallel cross (SPDC) designs [6].
lP

Previous studies [10–12] have proposed many statistical methods for diallel analysis. Among them, diallel

mating models proposed by Griffing [10] are the most widely used models for investigating the genetic
na

parameters (general and specific combining ability). For a partial diallel cross scheme, traditional studies [13]

used the ordinary least squares (OLS) to estimate GCA. However, if the diallel table is very sparse, the great
ur

reduction in the ratio of observed hybrids will be striking. Using the traditional OLS to evaluate many inbred lines

with a small sample of hybrids can’t ensure high accuracy.

Jo

Fortunately, with the advances of molecular biology, breeders can accurately understand the genetic structure of

breeding populations, and thus greatly improve the estimation of genetic parameters using genomic prediction

(GP). Estimated breeding values based on the genotypes of individuals were remarkably accurate [14, 15]. Some

studies [16, 17] used GS models for directly predicting agronomic traits in inbred lines, and some others [18–20]

used GS for predicting hybrid performance. Various methods, such as Genomic best linear unbiased prediction

(GBLUP), Bayes, the least absolute shrinkage and selection operator (LASSO) and machine learning [21–24]

have been developed for GP, and these differ with respect to assumptions about the marker effects. GBLUP and

RR-BLUP [21] assign identical variance to all loci and essentially treat all of them as equally important. In

BayesA, markers are assumed to have different variances and follows a posterior scaled inverse chi-square

distribution [22]. The prior in BayesB assumes the variance of markers to equal zero with probability π, and the
2
 Journal Pre-proof
complement with probability (1 – π) follows a scaled inverse chi-square distribution [22]. In Bayes Cπ, the

mixture probability π has a prior uniform distribution [25]. LASSO is a popular method for regression that uses a

penalty to achieve a sparse solution, and it is somewhat indifferent to closely correlated markers and tends to pick

one and ignore the others [26]. Machine learning is also an alternative for GS. It has been employed to enhance

the prediction of genetic values for wheat and maize [27, 28]. Although various models have been successfully

applied to GS, some studies [29–31] showed that not much variation in prediction accuracy among the different

models was observed.

In terms of the GCA analyses and hybrid prediction, Bernardo [32–34] was one of the first to advocate the

BLUP model [35] in maize. From then on, many studies have been reported to analyze GCA especially when

of
predicting hybrids. For predicting the GCA of a maize testcross population, Riedelsheimer et al. [2] proposed a

genomic selection (GS) approach based on RR-BLUP, showing that more efficient predictive procedures could be

ro
developed using genomic data. With a linear mixed model using the ASReml-R software [36], Kadam et al. [37]
-p
evaluated random inbreds derived from biparental families of maize. Using GBLUP and BayesB, Technow et al.

[38] investigated genome properties of the parental line based on the Dent × Flint heterotic pattern. Greenberg et
re
al. [6] developed a hierarchical Bayesian model to estimate quantitative genetic parameter from partial diallel
lP

cross designs. Werner et al. [39] considered GCA and specific combining ability (SCA) to apply RR-BLUP and

Bayesian models for predicting hybrid performance in oilseed rape using a collection of 220 paternal DH lines
na

and five male-sterile inbred lines. Recently, using GBLUP and a complete diallel cross design with twenty-eight

single-crosses formed between eight parental lines, Velez-Torres et al. [40] concluded that GS is a more effective
ur

and efficient approach to predict the GCA of maize lines compared with phenotyping method. However, in most
Jo

of the previous studies, more attention has been paid to the prediction accuracy of hybrids, and the prediction of

parental GCA is not the focus. For an SPDC maize population, few studies have been reported to systematically

investigate the GP accuracy for GCA.

The purpose of this study is to assess the efficacy of GP for estimating the GCA of maize inbred lines with

SPDC systems. As mentioned earlier, the accuracy of various GS models is similar. Considering that GBLUP is

more suitable for quantitative traits influenced by polygenes and its high computational efficiency [15], GBLUP

was adopted in this study. Using genome-wide SNPs called from a real maize data set of 266 inbred lines, genetic

and phenotypic values of all possible hybrids were simulated. And thus different hybrid sample sizes and different

distributions of parental lines involved in crossing were investigated. Such a scheme was implemented to assess

the efficacy of GP in estimating the GCA of lines. Then, the utility of statistical approaches was illustrated with an

example using an actual trait of maize. The methods we described would be useful in various sets of maize and
3
 Journal Pre-proof
other crops.

2. Materials and methods

2.1. Materials

2.1.1. Plant materials

The models were fitted to the maize data set from Yangzhou University. Partial diallel crossings between a total of

266 maize inbred lines were performed during the 2017 and 2018 maize growing seasons from field trials on the

experimental farms in Yangzhou and Taian, China, and two replicates were made in each environment. Regardless

of reciprocals, ear weight (EW) for 945 hybrids were collected to estimate the GCA of the inbred lines. The

of
phenotypes fitting the statistical model were the average performance of each hybrid from two environments. The

ro
266 inbred lines were genotyped, and 319,668 SNPs evenly distributed on chromosomes were called initially.

61,468 genome-wide SNPs were filtered by eliminating the heterozygosity >0.05 and the missing rate >0.05.
-p
Genotypes of the hybrids were inferred from SNPs of their inbred parents.
re
2.1.2. Simulations
lP

Based on the genotypes mentioned above, a large number of simulations were performed to assess the prediction

accuracy of different statistical approaches. Considering additive and dominant genetic effects of markers, several
na

traits of all the 35,245 possible hybrids with 200 QTLs and different heritabilities were randomly simulated. The

numbers of QTLs on chromosome 1–10 are 29, 32, 21, 25, 21, 12, 15, 21, 9, and 15, respectively. According to the
ur

study of Meuwissen et al. [22], the additive effects of the 200 QTLs were drawn from a gamma distribution with

shape parameter α = 0.4 and scale parameter β = 1.66. Half of the additive effects had positive effects and the
Jo

other half had negative effects. The dominant effects were determined as the product of the absolute additive

effects and the degree of dominance, which was drawn from a normal distribution with mean and variance equal

to 0.193 and 0.3122, respectively. In the 200 QTLs, two loci on chromosome 1 and chromosome 3 were found to

have weak over-dominant effects. For all the simulated hybrids, the ratio of dominant variance to additive

variance was 0.160. Normal independent error deviations with variances calculated were added to meet assumed

heritability, and the simulated phenotypes were centered and standardized to unit variance. Finally, three traits

with heritability of 0.7, 0.5, and 0.3, named T7, T5, and T3 were obtained.

2.1.3. Sampling designs

Two types of sampling designs were performed to evaluate the difference in predictive accuracy. First, for each

trait of the 35,245 hybrids, different numbers of hybrids (m = 500, 1000, and 2000) were sampled as the training
4
 Journal Pre-proof
sets to estimate the GCA of the 266 parental lines. Second, with a certain sample size of hybrids mentioned above,

three distributions of parental lines were designed. One was similar to a circulant diallel table. In this paper it was

called balanced sampling, which meant that all parental lines were involved in nearly an equal number of crosses

(designated s). Another was called random sampling, which meant that the crossing times of all the lines were

random but at least 1. The third was called unbalanced sampling, which meant that only part of the 266 lines (n =

200 or 150) were involved in random crossing. Each sampling method was randomly repeated 20 times to obtain

the average results of 20 replicates. The density plot for the crossing times of the 266 lines derived from the 20

samplings with different methods is shown in Fig. 1. Taking the first round of sampling with m = 500 as an

example, the detailed sampling scheme is illustrated in Table S1. In this study, for the random samplings, the

of
crossing times (s) of each inbred line was classified into three levels to compare the accuracy for line subsets with

different s value. About a quarter of the lines with the lowest s value were defined as low-frequency; a quarter of

ro
the lines with the highest s value were defined as high-frequency; the middle half was defined as
-p
medium-frequency.
re
lP
na
ur
Jo

5
 Journal Pre-proof

of
ro
-p
re
lP
na
ur
Jo

Fig. 1 – Density plot for the crossing times of the 266 inbred lines derived from 20 rounds of random, balanced and
unbalanced samplings, respectively.

2.2. Methods

The Griffing Model [10] was used for analyzing our SPDC schemes:

yij = μ + gi + gj + sij + εij

where yij is the phenotypic value of the hybrids between line i and line j (i, j = 1, 2, ..., n); μ is the overall mean;

gi and gj are the GCA effects of the ith parent and the jth parent, respectively; sij is the SCA effect for the cross

between the ith and jth parents; and 𝜀𝑖𝑗 is the random error effect.

Based on the above model, three statistical approaches were applied. One obtained estimations through the OLS
6
 Journal Pre-proof
and the other two utilized GP to predict the GCA. In our simulations, knowing the phenotypes of all the possible

hybrids, the true GCA of all the 266 inbred lines can be calculated with its definition. The coefficient of

determination (the squared Pearson correlation coefficient) between the true GCA and predicted GCA was

adopted to evaluate the accuracy of different statistical approaches.

2.2.1. OLS approach

In the matrix form, the vector for m observed hybrids can be represented by:

y = Xβ + ε

where y is an m × 1 vector, X is an m × (m + n + 1) matrix of coefficients, and β is an (m + n + 1) × 1 vector,

defined as ( μ, g1 , g2,…, gn , s1 , s2 ,…, sm )T which includes the overall mean, GCA and SCA effects. ε is an m × 1

of
random error vector. The restrictions ∑ni=1 gi =0 and ∑m
k=1 sk =0 were imposed on the combining ability effects by

ro
1…1 0
combining [ ] with the X matrix by rows, and combining (0, 0)T with y by columns. Thus, the OLS
0 1…1
1
-p
solutions could be achieved using the equation: β̂ =(X' X) (X' y). Because the diallel tables in our research were

very sparse, 𝑿′ 𝑿 was always a singular matrix. In this study, the Moore-Penrose generalized inverse was adopted
re
1
to calculate (X' X) using the R package MASS [41].
lP

2.2.2. GP approach
na

GBLUP is an efficient method using whole-genome markers to predict genetic values and phenotypes of interest.

It exploits the genomic relationships between training population and testing population to predict the genomic
ur

values for unknown individuals [42]. In this study, two GP approaches using GBLUP were performed for

predicting the GCA of the 266 lines. One was designated GP-I, which used the Griffing Model to directly estimate
Jo

the GCA. The other, designated GP-II, estimated the GCA by predicting the phenotypes of all possible hybrids.

The model of GP-I can be described as:

y = μ1 + Zg(1) g(1) + Zg(2) g(2) + Zs s + ε,

where y is an m × 1 vector for hybrid observations; μ is the overall mean; 1 is a vector of ones. For each hybrid,

g(1) and g(2) are vectors for the GCA effects of two parents, respectively; s is the vector for the SCA effects, and ε

is the vector for random residuals. Zg(1), Zg(2), Zs are incidence matrices related to g(1), g(2) and s. It is assumed that

g(1) ~ N(0, G(1) σ2(1) ), g(2) ~ N(0, G(2) σ2(2) ), s ~ N(0, Gs σ2s ), ε ~ N(0, Im σ2ε ), where σ2(1) and σ2(2) are the variances

of the GCA effects for two parents, respectively; σ2s is the variance of SCA effects; σ2ε is the residual variance;

I m is an m × m identity matrix; G(1) and G(2) are the additive genetic relationship matrices for two parents,

respectively; Gs is the non-additive genetic relationship matrix for the SCA effects. Using information from the
7
 Journal Pre-proof
genome-wide 61,836 SNP, G(1) , G(2) and Gs were defined as (𝑴(𝟏) 𝑴′(𝟏) )/𝑘(𝟏) , (𝑴(𝟐) 𝑴′(𝟐) )/𝑘(𝟐) and

(𝑴𝒔 𝑴′𝒔 )/𝑘𝒔, respectively. 𝑴(𝟏) and 𝑴(𝟐) are m × q matrices (q is the number of markers) for two parents which

specify genetic values at each locus (1, 0 and 1 for genotypes A1A1, A1A2 and A2A2, respectively). 𝑴𝒔 is an m ×

q matrix for hybrids which specifies dummy variables that take value 1 for heterozygous and 0 for homozygous

loci. k(1), k(2) and ks were calculated as trace (𝑴(𝟏) 𝑴′(𝟏) )/𝑚 , trace (𝑴(𝟐) 𝑴′(𝟐) )/𝑚 and trace (𝑴𝒔 𝑴′𝒔 )/𝑚 ,

respectively. Using GBLUP, for each observed hybrid, g(1) and g(2) were predicted. Then arithmetic mean of the

predicted GCA of each line was calculated.

The model of GP-II can be described as:

y = μ1 + Za a + Zd d + ε,

of
where y is an m × 1 vector for hybrid observations; μ is the overall mean; 1 is a vector of ones. a is the vector for

additive genetic effects of hybrids, d is the vector for dominance effects of hybrids. Za and Zd are incidence

ro
matrices related to a and d, respectively. It is assumed that a ~ N(0, Ga σ2a ), d ~ N(0, Gd σ2d ) and ε ~ N(0, Im σ2ε )
-p
where σ2a is the additive genetic variance; σ2d is the dominance variance; σ2ε is the residual variance; Ga and

Gd are the additive and dominance genetic relationship matrices, respectively. They were defined as (𝑴𝒂 𝑴′𝒂) )/
re
𝑘𝒂 and (𝑴𝒅 𝑴′𝒅)/𝑘𝒅, respectively. 𝑴𝒂 is an m × q matrix for hybrids which specifies genetic values (1, 0 and
lP

1 for genotypes A1A1, A1A2 and A2A2, respectively) at each locus. 𝑴𝒅 is an m × q matrix for hybrids which

specifies dummy variables that take value 1 for heterozygous and 0 for homozygous loci. ka and kd were calculated
na

as trace(𝑴𝒂 𝑴′𝒂) )/𝑚 and trace(𝑴𝒅 𝑴′𝒅)/𝑚, respectively. The proportions of the variances (σ2a , σ2d , σ2ε ) to the total

variance (𝜎 2 = σ2a + σ2d + σ2ε ) were defined as h2a = σ2a /𝜎 2 , h2d = σ2d /𝜎 2 and h2ε = σ2ε /𝜎 2 , respectively. Using
ur

GBLUP, for all the possible hybrids, a and d were predicted, and then the predicted phenotypes was derived from
Jo

the sum of a and d. Finally, the predicted GCA of each line could be calculated with the definition. The variance

ratios h2a , h2d and h2ε were estimated using restricted maximum likelihood (REML). In GP-I, the ratios of σ2(1)

and σ2(2) to the total variance were set equal to 0.5h2a, and the ratio of σ2s to the total variance was set equal to h2d .

The REML algorithm and the GBLUP model were performed using the R program [43].

3. Results

3.1. Comparison of the methods for estimating the GCA

For the traits T7, T5 and T3, based on random and balanced samplings, the prediction results of OLS, GP-I and

GP-II using different sample sizes were compared in Table 1. In each case, the trait T7 had the highest accuracy,

while the trait T3 had the lowest accuracy, showing that the estimation of GCA was largely dependent on

8
 Journal Pre-proof
heritability.

Table 1 – Comparison of accuracy using OLS, GP-I and GP-II with different sampling designs and training set sizes.
OLS GP-I GP-II
Sample
Trait Random Balanced Random Balanced Random Balanced
size
sampling sampling sampling sampling sampling sampling

T7 500 0.7361 0.7435 0.7709 0.7685 0.8068 0.8112

1000 0.8602 0.8711 0.8653 0.8740 0.8891 0.8995
2000 0.9364 0.9281 0.9340 0.9361 0.9443 0.9444
T5 500 0.5617 0.5758 0.6496 0.6477 0.6856 0.6926
1000 0.7432 0.7562 0.7790 0.7869 0.8068 0.8149
2000 0.8704 0.8679 0.8761 0.8776 0.8921 0.8960

of
T3 500 0.3634 0.3893 0.4659 0.4901 0.5185 0.5369
1000 0.5668 0.5773 0.6553 0.6551 0.6869 0.6862

ro
2000 0.7423 0.7434 0.7814 0.7902 0.8021 0.8112

-p
It is notable that the accuracy of each approach was always higher than the heritability of a target trait, showing

that the GCA for maize lines can be accurately predicted with SPDC designs. This character is very beneficial to
re
genetic improvement in breeding practice. For T7, derived from the random sampling i.e., when the sample size
lP

was 500, GP-II gave the highest accuracy (0.8068). It was 4.7% higher than that of GP-I (0.7709) and 9.6% higher

than that of OLS (0.7361); when the sample sizes were 1000 and 2000, GP-II gave the highest accuracy as before.
na

At this time, the accuracy of GP-I was only slightly higher than that of OLS and the advantage of GP-II was

smaller. For T5, when the sample sizes were 500 and 1000, the statistical approaches showed the similar pattern
ur

increasing from OLS to GP-II. When the sample size was 2000, the accuracy of GP-I was once again slightly
Jo

higher than that of OLS. For T3, no matter which sample size was adopted, GP-II gave much more accurate GCA

than GP-I and OLS, and the accuracies obtained by GP-I always substantially exceeded that of OLS.

Obviously, the sample size had a great influence on the accuracy. For all the three traits, statistical approaches

with the sample size of 2000 provided the highest accuracies, followed by those with the sample size of 1000,

reflecting that a big sample size could substantially contribute to the prediction. Additionally, the sample size

affected the significance of GP approaches over OLS. As mentioned above, the smaller the sample size, the higher

the level of significance was, showing that GP is particularly beneficial for the estimation of GCA with a sparse

partial diallel table. No matter which statistical approach was used, when comparing the random and balanced

samplings, their accuracies were not significantly different, showing that the random sampling commonly used in

breeding practice had little impact on the average accuracy of GCA.

In brief, GP-II performed the best in our research and would be a promising approach for estimating the GCA
9
 Journal Pre-proof
of maize and other crops.

3.2. Accuracy with different numbers of inbred lines involved in crossing

Sometimes, due to insufficient material resources or experimental budget, the number of inbred lines involved in

crossing is limited. Therefore, in this study, the effect of line quantity on prediction was explored. To simplify the

problem, only the GP-II approach that had the highest accuracy in the previous experiments was investigated. The

accuracy for the 266 lines with different numbers of lines (n = 150, 200, and 266) involved in random crossing

was plotted against the sampling number of hybrids (Fig. 2). As expected, for each of the three traits, the accuracy

was higher for the sampling designs with 266 lines over the sampling designs with 200 and 150 lines. For T7,

when the sample size was 500, 1000 and 2000, the highest accuracy obtained with n = 266 was 22.9%, 25.2%, and

of
30.4% higher than that with n = 150, respectively; For T5, when the sample size was 500, 1000 and 2000, the

ro
highest accuracy obtained with n = 266 was 20.8%, 22.3%, and 27.7% higher than that with n = 150, respectively;

For T3, when the sample size was 500, 1000 and 2000, the highest accuracy obtained with n = 266 was 14.0%,
-p
17.5%, and 24.4% higher than that with n = 150, respectively. It was clear that the advantage of more inbred lines
re
involved in crossing could be brought into full play when predicting a high-heritability trait with a big sample size

of hybrids.
lP
na
ur
Jo

Fig. 2 – Accuracy of GP-II for the 266 inbred lines with different numbers of lines (n = 150, 200, or 266) involved in the
random crossing

10
 Journal Pre-proof

3.3. Accuracy for subsets of inbred lines

The results of above research demonstrated that no significant differences were found between the accuracies of

the random and balanced samplings. However, for the random sampling, inbred lines involved in crossing with

different frequency must provide different amounts of information. Therefore, taking GP-II for instance, accuracy

for lines involved in crossing with different levels of frequency derived from the random sampling were

demonstrated in Table 2. It was shown that the lines involved in crossing with high-frequency always got the

highest accuracy, and those with low-frequency performed the worst. For predicting T3, with the sample size of

500, the high-frequency lines achieved the biggest increase (55.3%) in accuracy relative to the low-frequency

lines. That is to say, although the random sampling brings us similar accuracy to the balanced sampling, the

of
prediction for low-frequency lines may suffer from much lower accuracy, especially for a low-heritability trait

with a small sample size of hybrids.

ro
-p
Table 2 – Accuracy of GP-II for inbred lines involved in crossing with different levels of frequency.

Sample Lines involved in crossing Lines involved in crossing Lines involved in crossing
Trait
re
size with low-frequency with medium-frequency with high-frequency

T7 500 0.7304 0.8041 0.8675

lP

1000 0.8372 0.8932 0.9285

2000 0.9242 0.9463 0.9592
T5 500 0.5882 0.6730 0.7802
na

1000 0.7173 0.8196 0.8530

2000 0.8649 0.8942 0.9129
T3 500 0.4081 0.4954 0.6337
ur

1000 0.6087 0.6870 0.7581

Jo

Unbalanced sampling design was also concerned in this research. The accuracies of GP-II for lines involved

and not involved in crossing are summarized in Table 3. As expected, the accuracies were always much higher for

the involved lines over the non-involved lines. In most cases, the accuracy increases with the sample size. It is

interesting to note that the accuracy of non-involved lines is always less sensitive to the sample size than that of

the involved lines. In particular for T7, the accuracy of the 166 non-involved lines with m = 2000 (0.4105) was

6.4% higher than that with m = 500 (0.3858), while the advantage of the 150 involved lines was 10.5%; in the

same way, the accuracy of the 66 non-involved lines with m = 2000 (0.4594) was only 4.3% higher than that with

m =500 (0.4405), while the advantage of the 200 involved lines was 14.5%. For T5 and T3, the pattern of the

accuracy was similar to that for T7.

11
 Journal Pre-proof
Table 3 – Accuracy of GP-II for inbred lines involved and not involved in crossing derived from the unbalanced samplings

Number of lines Number of lines not Sample Accuracy for Accuracy for
Trait
involved in crossing involved in crossing size involved lines non-involved lines
T7 150 116 500 0.8755 0.3858
1000 0.9359 0.4143
2000 0.9676 0.4105
200 66 500 0.8359 0.4405
1000 0.9182 0.4578
2000 0.9570 0.4594

T5 150 116 500 0.7511 0.3405

1000 0.8746 0.3790
2000 0.9359 0.3925

of
200 66 500 0.7166 0.3869
1000 0.8337 0.3978

ro
2000 0.9164 0.4454

T3 150 116 500 0.6002 0.2745

-p
1000 0.7662 0.3486
2000 0.8699 0.3564
re
200 66 500 0.5549 0.2689
1000 0.7081 0.3416
lP

2000 0.8349 0.4007

na

In each case, accuracy of involved lines with n = 200 was lower than that with n = 150. A good interpretation

was that the involved lines with n = 150 has higher frequency in crossing, which showed again the contribution of
ur

high-frequency to prediction. On the contrary, accuracies of non-involved lines with n = 200 were almost higher

than that with n = 150. The reason may be that the advantage of high-frequency cannot work on the non-involved
Jo

lines, and a sample of hybrids derived from more lines can benefit the prediction.

3.4. Prediction results for EW of maize

In addition to the simulated studies, actual EW of 945 hybrids were used to estimate the GCA of the 266 inbred

lines. The predicted GCA values were sorted in descending order and the prediction results of the top 20 inbred

lines are demonstrated in Table 4. As shown, eight common varieties, including B93, B57, B108, B74, B214,

B275, B254 and B167 were screen out by all the three statistical approaches. It is noteworthy that the coefficient

of determination between the predicted GCA of GP-II and OLS was the highest (0.8703), and up to sixteen

common varieties were selected by the two approaches simultaneously, indicating that the intersection between

the results of GP-II and OLS was more reliable in this scenario. Moreover, the absolute GCA values predicted by

12
 Journal Pre-proof
GP-I were lower than those predicted by OLS and GP-II. The reason may be that the non-additive effects were

excluded from the estimates of GCA when using GP-I. Although this exclusion may not affect the predictive

accuracy, the loss of non-additive variance will inevitably reduce the absolute values of the predicted GCA.

Table 4 – The top 20 inbred lines of the GCA for EW using OLS, GP-I, and GP-II.
OLS GP-I GP-II

Top 20 Serial Estimated Serial Estimated Serial Estimated

number GCA number GCA number GCA

1 B093 193.7 B254 152.3 B057 183.4

2 B079 185.9 B275 148.6 B068 144.4

of
3 B057 180.2 B057 136.3 B093 139.5
4 B108 171.7 B218 125.5 B131 136.0

ro
5 B074 170.9 B214 122.0 B108 134.9
6 B053 160.4 B241 121.4 B003 133.1
7 B106 147.2 B263 118.4 B016 130.7
-p
8 B214 146.7 B271 118.4 B052 129.8
9 B052 146.0 B167 118.3 B074 129.4
re
10 B275 140.6 B093 116.5 B079 129.0
11 B063 138.8 B264 115.9 B117 126.7
lP

12 B254 137.7 B211 114.9 B053 125.0

13 B047 136.2 B208 109.4 B167 123.0
14 B167 135.3 B226 106.9 B028 121.8
na

15 B010 128.4 B117 106.7 B006 121.0

16 B016 125.9 B233 99.4 B047 120.0
17 B003 124.9 B074 98.6 B254 118.6
ur

18 B071 124.7 B270 96.7 B275 116.4

19 B037 124.7 B108 96.4 B106 115.4
Jo

20 B068 123.1 B276 95.8 B214 114.3

4. Discussion
With a complete diallel cross scheme, the GCA of inbred line can be easily calculated with its definition [1].

However, because of the large number of possible crosses, Kempthorne and Curnow [7] suggested the partial

diallel cross to evaluate inbred lines. Each of the n lines are crossed with s other lines, and there will be ns/2

crosses in the whole set. In this scenario, although the OLS was widely used for estimating the GCA [13, 44],

genomic information was ignored. GS uses all molecular markers for predicting the performance of the candidates,

and it has shown tangible genetic gains in maize breeding [14]. Recently, Alves et al. [45] pointed out that GS can

be used to estimate genetic parameters accurately in maize hybrids. For maize inbred lines, there is no reason to
13
 Journal Pre-proof
doubt the advantage of GS over the phenotype-based OLS. Comparisons of the three approaches in the present

study have shown that GP-II and GP-I are superior to OLS, showing that prediction with genomic data can help

improve the estimation of GCA for inbred lines based on SPDC designs.

Riedelsheimer et al. [2] crossed 285 diverse Dent inbred lines with two testers and predicted the GCA using

genomic and metabolic information. Predictive accuracies (the Pearson correlation) ranged from 0.72 to 0.81 for

GP, which are similar to ours for T3 and T5. However, a fivefold cross-validation scheme was applied in their

prediction and the predicted GCA values of one subset were estimated using the observed GCA values of the other

four subsets. In our research, the situation was quite different. The partial diallel table was assumed to be very

sparse. Even if the hybrid sample size equals 2000, the ratio of the sample size (2000) to the possible hybrid size

of
(35,245) is only 5.7%, making it impossible to obtain a GCA observation of any line. In other words, our

prediction is based on the SPDC. The sampling hybrids are treated as the training set, and their number is often

ro
lager than the number of lines with identified GCA. Maybe this is the reason why our accuracies are slightly
-p
higher than those of Riedelsheimer et al. [2].

Population design plays a vital role in breeding programs, and partial diallel cross is preferable in many cases.
re
For instance, Miranda Filho and Vencovsky [13] estimated the GCA for ear length of maize in a partial diallel
lP

cross with n = 10 and s = 3. Reis et al. [44] estimated the genetic parameters using a partial circulant diallel cross

design with n = 34 (two groups of parents) and different sizes of s (from 2 to 5). Analysis using cross-validation
na

process showed that the accuracy increased as the value of s increased. Our balanced and unbalanced sampling

designs just mimicked the partial diallel cross scheme. With n = 266, the hybrid sample size was set to 500, 1000,
ur

and 2000, respectively. Correspondingly, the mean of s was 3.8, 7.5, and 15.0. Note that s was much less than n 

1. The inadequate phenotypic information of each parental line in crossing couldn’t guarantee the accurate
Jo

estimation of GCA. Although Vivas et al. [9] declared that it is possible to obtain good agreement (correlation

coefficient above 0.8) with s = 3, our accuracies with s = 3.8 are still lower than those with s = 7.5 and s = 15.0.

When evaluating the efficiency of the circulant diallel, Veiga et al. [46] pointed out that it is advantageous to

increase the s value for a low-heritability trait. In our research, with the sample size of 500 for predicting T3,

accuracies of OLS were substantially lower than those of GP-I and GP-II, showing that the reduction in the s

value decreased the potential accuracy of OLS. However, it is gratifying that the GP approaches using genomic

information have been demonstrated to partly compensate the “small s” problem.

The unbalanced sampling design was also worthy of attention in breeding practice. Because of experimental

cost and complex factors in field trails, involving all inbred lines in crossing is impossible. Previous studies [19,

47] have adopted the strategy of predicting untested single-cross hybrids in maize and rice. However, few GS
14
 Journal Pre-proof
studies have been undertaken in predicting the GCA of the lines that never participate in crossing. In this case, the

traditional phenotype-based OLS is impracticable. Our research has demonstrated that the GCA of the

non-involved lines could also be estimated using GP approaches based on SPDC designs. This strategy allows a

reliable selection of more inbred lines for their potential to create superior hybrids. But on the other hand, lines

never or seldom involved in crossing were found to have lower accuracy than the involved lines. In this respect,

our results are in agreement with those reported by Fristche-Neto et al. [48] who have indicated that the number of

parents and the crosses per parent in the training sets should be maximized when predicting maize hybrid

performance.

Empirical studies [31, 49, 50] have shown that heritability of target traits and training population size are two

of
important factors affecting the accuracy. In our research, the accuracies always increase with the heritability and

the sample size, which was in agreement with previous theoretical results. Reduction of the error effects may also

ro
help increase the accuracy. A previous study [43] has shown that the trouble of predicting a low-heritability trait is
-p
the significant standard deviation caused by errors. In the present study, for predicting the GCA, the accuracies of

GP-II were not only higher than the heritability of target traits, but also higher than the accuracies of hybrid
re
prediction (Table 5), indicating that the prediction of GCA may be more effective than that of hybrids. According
lP

to the study of Daetwyler et al. [50], the Pearson correlation coefficient between the true breeding values and
𝑁𝑝 ℎ2
estimated breeding values (𝑟𝑔𝑔̂ ) is a function of heritability (h2). They derived the equation: 𝑟𝑔𝑔̂ = √𝑁 2 +𝑀 ,
𝑝ℎ
na

where Np is the number of individuals, and Me is the number of independent chromosome segments. Furthermore,

Resende et al. [51] pointed out that the Pearson correlation coefficient between the true phenotypic values and
ur

estimated phenotypic values of GS can be expressed as 𝑟𝑦𝑦̂ = 𝑟𝑔𝑔̂ ℎ. Theoretically, 𝑟𝑔𝑔̂ is less than 1, and 𝑟𝑦𝑦̂ is
Jo

less than h. In our manuscript, the accuracy is the squared Pearson correlation coefficient between the true GCA

and predicted GCA. By definition, GCA is essentially the average performance of a line in hybrid combinations,

and the average of predicted genetic values of the possible hybrids greatly eliminated the influence of errors. This
2 2
factor may make the accuracy between 𝑟𝑦𝑦̂ and 𝑟𝑔𝑔̂ , so the accuracy in our manuscript was higher than the

heritability of the trait. This result agrees well with the fact that the accuracy of GS excluding error effects can be

much higher than the predictive ability containing error effects [31, 50].

15
 Journal Pre-proof
Table 5 – Accuracy of GP-II for the GCA of 266 inbred lines and for all the potential hybrids with the random samplings.

Trait Sample size Accuracy for GCA Accuracy for hybrids

T7 500 0.8068 0.7754

1000 0.8891 0.8558
2000 0.9443 0.9106
T5 500 0.6856 0.6555
1000 0.8068 0.7742
2000 0.8921 0.8560
T3 500 0.5185 0.4955
1000 0.6869 0.6536
2000 0.8021 0.7594

Maize breeding involves two critical steps, developing superior inbred lines from breeding populations and

of
identifying elite combinations of two inbred lines [52]. With the development of DH and other technologies,

ro
breeders have been able to develop a large number of inbred lines which need to be evaluated by their
-p
performance in crosses. However, the number of potential crosses grows rapidly, making the field evaluation of

hybrid performance time and resource intensive. Sparse partial diallel tables are becoming more and more
re
common in breeding practice. GCA is mainly a measure of additive effects, and it is in response to selection of
lP

inbred lines. Accurate prediction of the GCA will enhance the efficiency of inbred line selection, and then

accelerate the hybrid breeding. In actual breeding projects, especially for the scenario with only datasets based on
na

SPDC designs, breeders can evaluate their inbred lines using the methods proposed in the present study. Then, for

different heterotic groups, top inbred lines can be selected, and a few corresponding testers can be used to perform
ur

the validation by field trials. We believe that in this way, the efficiency and accuracy of maize breeding can be

improved.
Jo

Previous studies [3, 53, 54] have used maize introgression lines or recombination inbred lines to perform

testcrosses for detecting significant loci of GCA. In all these experiments, phenotypes of hybrids were observed

and the true GCA value could be calculated with certainty. However, in most breeding programs, only a small part

of possible hybrids can be identified in field trials, and thus the detection work can’t be carried out. In such cases,

methods proposed in the present study guarantee a reliable estimation for the GCA, providing an opportunity for

further detection of significant loci. This strategy may open up a promising research direction for inbred line

selection in maize and other crops.

Declaration of competing interest

The authors declare that they have no conflicts of interest.

16
 Journal Pre-proof

Acknowledgments
This work was supported by the National Key Research and Development Program of China (2016YFD0100303),

the National Natural Science Foundation of China (31801028, 31902101), the Open Research Fund of State Key

Laboratory of Hybrid Rice (Wuhan University) (KF201701) and the Priority Academic Program Development of

Jiangsu Higher Education Institutions.

Author contributions
Xin Wang performed the analysis and wrote the paper. Zhenliang Zhang, Yang Xu, and Pengchen Li collected the

data. Xuecai Zhang assisted with the analysis. Chenwu Xu conceived and designed the analysis.

Appendix A. Supplementary data

of
Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.xx.xxx.

ro
References
[1] G.F. Sprague, L.A. Tatum, General vs. specific combining ability in single crosses of corn, Agron. J. 34 (1942) 923–932.
[2]
-p
C. Riedelsheimer, A. Czedik-Eysenberg, C. Grieder, J. Lisec, F. Technow, R. Sulpice, T. Altmann, M. Stitt, L. Willmitzer,
A.E. Melchinger, Genomic and metabolic prediction of complex heterotic traits in hybrid maize, Nat. Genet. 44 (2012)
re
217–220.
[3] J. Huang, H. Qi, X. Feng, Y. Huang, L. Zhu, B. Yue, General combining ability of most yield-related traits had a genetic
lP

basis different from their corresponding traits per se in a set of maize introgression lines, Genetica 141 (2013) 453–461.
[4] H.A. Sadalla, M.O. Barznji, S.A. Kakarash, Full diallel crosses for estimation of genetic parameters in maize, Iraqi J. Agric.
Sci. 48 (2017) 30–40.
na

[5] X.M. Fan, Y.D. Zhang, D.P. Jeffers, Y.Q. Bi, M.S. Kang, X.F. Yin, Combining ability of yellow lines derived from
CIMMYT populations for use in subtropical and tropical midaltitude maize production environments, Crop Sci. 58 (2018)
169–179.
ur

[6] A.J. Greenberg, S.R. Hackett, L.G. Harshman, A.G. Clark, A hierarchical Bayesian model for a novel sparse partial diallel
crossing design, Genetics 185 (2010) 361–373.
Jo

[7] O. Kempthorne, R. Curnow, The partial diallel cross, Biometrics 17 (1961) 229–250.
[8] O. Kempthorne, A class of experimental designs using blocks of two plots, Ann. Math. Statist. (1953) 76–84.
[9] M. Vivas, S.F. Silveira, A.P. Viana, A.T. Amaral, D.L. Cardoso, M.G. Pereira, Efficiency of circulant diallels via mixed
models in the selection of papaya genotypes resistant to foliar fungal diseases, Gen. Mol. Res. 13 (2014) 4797–4804.
[10] B. Griffing, Concept of general and specific combining ability in relation to diallel crossing systems, Aust. J. Biol. Sci. 9
(1956) 463–493.
[11] J. Jinks, B. Hayman, The theory and analysis of diallel crosses, Genetics 43 (1953) 63–85.
[12] C. Gardner, S. Eberhart, Analysis and interpretation of the variety cross diallel and related populations, Biometrics (1966)
439–452.
[13] J.B. Miranda Filho, R. Vencovsky, The partial circulant diallel cross at the interpopulation level, Genet. Mol. Biol. 22 (1999)
249–255.
[14] J. Crossa, P. Perez-Rodriguez, J. Cuevas, O. Montesinos-Lopez, D. Jarquin, G. de los Campos, J. Burgueno, J.M.
Gonzalez-Camacho, S. Perez-Elizalde, Y. Beyene, S. Dreisigacker, R. Singh, X.C. Zhang, M. Gowda, M. Roorkiwal, J.
Rutkoski, R.K. Varshney, Genomic selection in plant breeding: methods, models, and perspectives, Trends Plant Sci. 22
(2017) 961–975.
[15] X. Wang, Y. Xu, Z. Hu, C. Xu, Genomic selection methods for crop improvement: current status and prospects, Crop J. 6
17
 Journal Pre-proof
(2018) 330–340.
[16] J. Crossa, G. de los Campos, P. Perez, D. Gianola, J. Burgueno, J. Luis Araus, D. Makumbi, R.P. Singh, S. Dreisigacker, J.
Yan, V. Arief, M. Banziger, H.-J. Braun, Prediction of genetic values of quantitative traits in plant breeding using pedigree
and molecular markers, Genetics 186 (2010) 713–724.
[17] J. Crossa, Y. Beyene, S. Kassa, P. Pérez, J.M. Hickey, C. Chen, G. de los Campos, J. Burgueño, V.S. Windhausen, E.
Buckler, Genomic prediction in maize breeding populations with genotyping-by-sequencing, G3-Genes Genomics Genet. 3
(2013) 1903–1926.
[18] F. Technow, C. Riedelsheimer, T.A. Schrag, A.E. Melchinger, Genomic prediction of hybrid performance in maize with
models incorporating dominance and population specific marker effects, Theor. Appl. Genet. 125 (2012) 1181–1194.
[19] S. Xu, D. Zhu, Q. Zhang, Predicting hybrid performance in rice using genomic best linear unbiased prediction, Proc. Natl.
Acad. Sci. U. S. A. 111 (2014) 12456–12461.
[20] X. Wang, Y. Xu, P. Li, M. Liu, C. Xu, Z. Hu, Efficiency of linear selection index in predicting rice hybrid performance,
Mol. Breed. 39 (2019) 77.
[21] P.M. VanRaden, Efficient methods to compute genomic predictions, J. Dairy Sci. 91 (2008) 4414–4423.

of
[22] T.H. Meuwissen, B.J. Hayes, M.E. Goddard, Prediction of total genetic value using genome-wide dense marker maps,
Genetics 157 (2001) 1819–1829.

ro
[23] R. Tibshirani, Regression shrinkage and selection via the lasso, J. R. Stat. Soc. B (1996) 267–288.
[24] O. González-Recio, G.J. Rosa, D. Gianola, Machine learning methods and predictive ability metrics for genome-wide
-p
prediction of complex traits, Livest. Sci. 166 (2014) 217–231.
[25] D. Habier, R.L. Fernando, K. Kizilkaya, D.J. Garrick, Extension of the Bayesian alphabet for genomic selection, BMC
re
Bioinformatics 12 (2011) 186.
[26] J. Friedman, T. Hastie, R. Tibshirani, Regularization paths for generalized linear models via coordinate descent, J. Stat.
Software 33 (2010) 1–22.
lP

[27] G. de Los Campos, D. Gianola, G.J. Rosa, K.A. Weigel, J. Crossa, Semi-parametric genomic-enabled prediction of genetic
values using reproducing kernel Hilbert spaces methods, Genet. Res. 92 (2010) 295–308.
[28] J. González-Camacho, G. de Los Campos, P. Pérez, D. Gianola, J. Cairns, G. Mahuku, R. Babu, J. Crossa,
na

Genome-enabled prediction of genetic values using radial basis function neural networks, Theor. Appl. Genet. 125 (2012)
759–771.
ur

[29] G. Moser, B. Tier, R.E. Crump, M.S. Khatkar, H.W. Raadsma, A comparison of five methods to predict genomic breeding
values of dairy bulls from genome-wide SNP markers, Genet. Sel. Evol. 41 (2009) 56.
[30] H.H. Neves, R. Carvalheiro, S.A. Queiroz, A comparison of statistical methods for genomic selection in a mice population,
Jo

BMC Genet. 13 (2012) 100.

[31] X. Wang, Z. Yang, C. Xu, A comparison of genomic selection methods for breeding value prediction, Sci. Bull. 60 (2015)
925–935.
[32] R. Bernardo, Prediction of maize single-cross performance using RFLPs and information from related hybrids, Crop Sci.
34 (1994) 20–25.
[33] R. Bernardo, Genetic models for predicting maize single-cross performance in unbalanced yield trial data, Crop Sci. 35
(1995) 141–147.
[34] R. Bernardo, Best linear unbiased prediction of maize single-cross performance, Crop Sci. 36 (1996) 50–56.
[35] C.R. Henderson, Best linear unbiased estimation and prediction under a selection model, Biometrics 31 (1975) 423–447.
[36] D.G. Butler, B.R. Cullis, A.R. Gilmour, B.J. Gogel. Mixed models for S language environments: ASReml-R reference
manual, Queensland Department of Primary Industries and Fisheries, Australia, 2009.
[37] D.C. Kadam, S.M. Potts, M.O. Bohn, A.E. Lipka, A.J. Lorenz, Genomic prediction of single crosses in the early stages of a
maize hybrid breeding pipeline, G3-Genes Genomics Genet. 6 (2016) 3443–3453.
[38] F. Technow, T.A. Schrag, W. Schipprack, E. Bauer, H. Simianer, A.E. Melchinger, Genome properties and prospects of
genomic prediction of hybrid performance in a breeding program of maize, Genetics 197 (2014) 1343–1355.

18
 Journal Pre-proof
[39] C.R. Werner, L. Qian, K.P. Voss-Fels, A. Abbadi, G. Leckband, M. Frisch, R.J. Snowdon, Genome-wide regression models
considering general and specific combining ability predict hybrid performance in oilseed rape with similar accuracy
regardless of trait architecture, Theor. Appl. Genet. 131 (2018) 299–317.
[40] M. Velez-Torres, J.J. Garcia-Zavala, M. Hernandez-Rodriguez, R. Lobato-Ortiz, J.J. Lopez-Reynoso, I. Benitez-Riquelme,
J.A. Mejia-Contreras, G. Esquivel-Esquivel, J.D. Molina-Galan, P. Perez-Rodriguez, X.C. Zhang, Genomic prediction of
the general combining ability of maize lines (Zea mays L.) and the performance of their single crosses, Plant Breed. 137
(2018) 379–387.
[41] W.N. Venables, B.D. Ripley, Modern applied statistics with S. 4th ed, Springer, New York, USA, 2002.
[42] Y. Xu, X. Wang, X. Ding, X. Zheng, Z. Yang, C. Xu, Z. Hu, Genomic selection of agronomic traits in hybrid rice using an
NCII population, Rice 11 (2018) 32.
[43] X. Wang, L. Li, Z. Yang, X. Zheng, S. Yu, C. Xu, Z. Hu, Predicting rice hybrid performance using univariate and
multivariate GBLUP models based on North Carolina mating design II, Heredity 118 (2017) 302–310.
[44] A.J.d.S. Reis, L.J. Chaves, J.B. Duarte, E.M. Brasil, Prediction of hybrid means from a partial circulant diallel table using
the ordinary least square and the mixed model methods, Genet. Mol. Biol. 28 (2005) 314–320.

of
[45] F. Alves, A. Granato, G. Galli, D. Lyra, R. Fritsche-Neto, G. de los Campos, Bayesian analysis and prediction of hybrid
performance, Plant Methods 15 (2019) 14.

ro
[46] R.D. Veiga, D.F. Ferreira, M.A.P. Ramalho, Efficiency of circulant diallels in parental choice, Pesq. Agropec. Bras. 35
(2000) 1395–1406. -p
[47] K. Dias, H. Piepho, L. Guimarães, P. Guimarães, S. Parentoni, M. Pinto, R. Noda, J. Magalhães, C. Guimarães, A. Garcia,
Novel strategies for genomic prediction of untested single-cross maize hybrids using unbalanced historical data, Theor.
re
Appl. Genet. (2019) 1–13.
[48] R. Fristche-Neto, D. Akdemir, J.L. Jannink, Accuracy of genomic selection to predict maize single-crosses obtained
through different mating designs, Theor. Appl. Genet. 131 (2018) 1153–1162.
lP

[49] H. Zhang, L. Yin, M. Wang, X. Yuan, X. Liu, Factors affecting the accuracy of genomic selection for agricultural economic
traits in maize, cattle, and pig populations, Front. Genet. 10 (2019) 189.
[50] H.D. Daetwyler, R. Pong-Wong, B. Villanueva, J.A. Woolliams, The impact of genetic architecture on genome-wide
na

evaluation methods, Genetics 185 (2010) 1021–1031.

[51] M.F. Resende, P. Muñoz, M.D. Resende, D.J. Garrick, R.L. Fernando, J.M. Davis, E.J. Jokela, T.A. Martin, G.F. Peter, M.
ur

Kirst, Accuracy of genomic selection methods in a standard data set of loblolly pine (Pinus taeda L.), Genetics 190 (2012)
1503–1510.
[52] T. Guo, H. Li, J. Yan, J. Tang, J. Li, Z. Zhang, L. Zhang, J. Wang, Performance prediction of F 1 hybrids between
Jo

recombinant inbred lines derived from two elite maize inbred lines, Theor. Appl. Genet 126 (2013) 189–201.
[53] H. Qi, J. Huang, Q. Zheng, Y. Huang, R. Shao, L. Zhu, Z. Zhang, F. Qiu, G. Zhou, Y. Zheng, Identification of combining
ability loci for five yield-related traits in maize using a set of testcrosses with introgression lines, Theor. Appl. Genet. 126
(2013) 369–377.
[54] Z.Q. Zhou, C.S. Zhang, X.H. Lu, L.W. Wang, Z.F. Hao, M.S. Li, D.G. Zhang, H.J. Yong, H.Y. Zhu, J.F. Weng, X.H. Li,
Dissecting the genetic basis underlying combining ability of plant height related traits in maize, Front. Plant Sci. 9 (2018)
1117.

19