Theoretical and Applied Genetics
When less can be better: How can we make genomic selection more
cost‑eective and accurate in barley?
· Paulino Pérez‑Rodríguez
· José Crossa
· François Belzile
Received: 13 November 2017 / Accepted: 24 May 2018
© Springer-Verlag GmbH Germany, part of Springer Nature 2018
Key message We were able to obtain good prediction accuracy in genomic selection with ~ 2000 GBS-derived SNPs.
SNPs in genic regions did not improve prediction accuracy compared to SNPs in intergenic regions.
Abstract Since genotyping can represent an important cost in genomic selection, it is important to minimize it without
compromising the accuracy of predictions. The objectives of the present study were to explore how a decrease in the unit
cost of genotyping impacted: (1) the number of single nucleotide polymorphism (SNP) markers; (2) the accuracy of the
resulting genotypic data; (3) the extent of coverage on both physical and genetic maps; and (4) the prediction accuracy (PA)
for six important traits in barley. Variations on the genotyping by sequencing protocol were used to generate 16 SNP sets
ranging from ~ 500 to ~ 35,000 SNPs. The accuracy of SNP genotypes ﬂuctuated between 95 and 99%. Marker distribution
on the physical map was highly skewed toward the terminal regions, whereas a fairly uniform coverage of the genetic map
was achieved with all but the smallest set of SNPs. We estimated the PA using three statistical models capturing (or not)
the epistatic eﬀect; the one modeling both additivity and epistasis was selected as the best model. The PA obtained with
the diﬀerent SNP sets was measured and found to remain stable, except with the smallest set, where a signiﬁcant decrease
was observed. Finally, we examined if the localization of SNP loci (genic vs. intergenic) aﬀected the PA. No gain in PA was
observed using SNPs located in genic regions. In summary, we found that there is considerable scope for decreasing the cost
of genotyping in barley (to capture ~ 2000 SNPs) without loss of PA.
Three selection strategies are currently practiced in the ﬁeld
of plant breeding: phenotypic selection, marker-assisted and
genomic selection (Ortiz Ríos 2015). The impetus for devel-
oping new breeding procedures has always been the desire
to make breeding more eﬃcient, quicker and less costly.
During centuries, plant breeding was mainly achieved by
crossing parents with the desired traits to generate genetic
variation through recombination, and then selecting the best
segregating oﬀspring based on phenotypic evaluation via
extensive ﬁeld and/or greenhouse trials throughout genera-
tions, across locations, and over time (Walsh 2001; Cattivelli
et al. 2011; Ortiz Ríos 2015). Phenotypic selection tends
to achieve relatively lower genetic gain for complex traits
with low heritability compared with traits with high herit-
ability (Bhat et al. 2016; Rajsic et al. 2016). For complex
traits, this process can be extremely time-consuming (Sallam
and Smith 2016) and the logistics of implementing can be
resource-intensive endeavor and very costly (Spindel et al.
Communicated by Marcos Malosetti.
Electronic supplementary material The online version of this
article (https ://doi.org/10.1007/s0012 2-018-3120-8) contains
supplementary material, which is available to authorized users.
* François Belzile
Département de Phytologie, Université Laval, Quebec City,
Institut de Biologie Intégrative et des Systèmes (IBIS),
Université Laval, Quebec City, QC, Canada
Programa de Estadística y Cómputo, Colegio de
Postgraduados, CP 56230 Montecillos, Edo. de México,
Biometrics and Statistics Unit, International Maize
and Wheat Improvement Center (CIMMYT), Apdo. Postal
6-641, 06600 Mexico City, Mexico