A new genomic tool for walnut (Juglans regia L.): development and validation of the high‐density Axiom™ J. regia 700K SNP genotyping array

Abstract

Summary Over the last 20 years, global production of Persian walnut ( Juglans regia L.) has grown enormously, likely reflecting increased consumption due to its numerous benefits to human health. However, advances in genome‐wide association ( GWA ) studies and genomic selection ( GS ) for agronomically important traits in walnut remain limited due to the lack of powerful genomic tools. Here, we present the development and validation of a high‐density 700K single nucleotide polymorphism ( SNP ) array in Persian walnut. Over 609K high‐quality SNP s have been thoroughly selected from a set of 9.6 m genome‐wide variants, previously identified from the high‐depth re‐sequencing of 27 founders of the Walnut Improvement Program ( WIP ) of University of California, Davis. To validate the effectiveness of the array, we genotyped a collection of 1284 walnut trees, including 1167 progeny of 48 WIP families and 26 walnut cultivars. More than half of the SNP s (55.7%) fell in the highest quality class of ‘ Poly High Resolution ’ ( PHR ) polymorphisms, which were used to assess the WIP pedigree integrity. We identified 151 new parent‐offspring relationships, all confirmed with the Mendelian inheritance test. In addition, we explored the genetic variability among cultivars of different origin, revealing how the varieties from Europe and California were differentiated from Asian accessions. Both the reconstruction of the WIP pedigree and population structure analysis confirmed the effectiveness of the Applied Biosystems™ Axiom™ J. regia 700K SNP array, which initiates a novel genomic and advanced phase in walnut genetics and breeding.

STAGdb: a 30K SNP genotyping array and Science Gateway for Acropora corals and their dinoflagellate symbionts

“…636 This was evident in the different percent heterozygous estimate of the fixed probes in the 637 between laboratory replicate extractions (Table S10). However, our technical error is similar to 638 previous genotype concordance estimates ranging from 0.2% to 2.4% for replicates of a given 639 subject genotyped on Affymetrix SNP arrays for humans (Hong et al 2012), rainbow trout (Palti 640 et al 2015), soybean (Lee et al 2015) and walnut (Marrano et al 2019). In that latter study, the 641 variation was also higher between technical replicates than biological replicates, which the 642 authors attributed to DNA quality.…”

High-quality chromosome-scale assembly of the walnut (Juglans regiaL) reference genome

“…Serr (JrSerr_v1.0; 534.7 Mb) (Zhu et al 2019) can be explained by structural variation (e.g., copy number and presence/absence variants), whose central role in explaining intraspecific genomic and phenotypic diversity has been reported in different plant species (Springer et al 2009; Marroni et al 2014). In addition, the higher number of unanchored scaffolds in Chandler v2.0 compared to JrSerr_v1.0 can represent autozygous genomic regions of Chandler, devoid of segregating markers and, therefore, difficult to anchor to linkage genetic maps (Luo et al 2015), as also suggested by the higher fixation index ( F ) observed in Chandler (0.03) than Serr (-0.29) in previous genetic surveys (Marrano et al 2018).…”

“…In these nine low heterozygous regions, we found 1,536 SNPs in total ( Figure 2 ), of which only 25 were tiled on the Axiom J. regia 700K SNPs array. The absence of these polymorphisms segregating in Chandler in the SNP array could be related to either a failed identification during the SNP calling due to the highly fragmented reference genome v1.0 or with the SNP exclusion during the filtering process applied to build the genotyping array (Marrano et al 2018). The low number of Chandler heterozygous SNPs in the array affected the end of Chr15 the most, causing a reduction in the genetic length of the corresponding linkage group ( Figure 1 ), as well as leaving unexplored 4 Mb of Chandler genetic variability, which is now accessible thanks to the new chromosome-scale reference genome.…”

“…Considerable phenotypic and genetic variability can be observed in this wide distribution area, especially in the Eastern countries, where walnuts can still be found in wild fruit forests. Many studies on genetic diversity in walnut have outlined a genetic differentiation between Eastern and Western genotypes (Ebrahimi et al 2016; Marrano et al 2018). Moreover, walnuts from Eastern Europe, Central Asia, and China exhibit higher genetic diversity and a higher number of rare alleles than the genotypes from Western countries (Bernard et al 2018a).…”

“…For the first time, it was possible to explore the gene space of Persian walnut with the prediction of 32,498 gene models, providing the basis to untangle complex phenotypic pathways, such as those responsible for the synthesis of phenolic compounds. The availability of a reference genome marked the beginning of a genomics phase in Persian walnut, allowing whole-genome resequencing (Stevens et al 2018; Zhang et al 2019), the development of high-density genotyping tools (Marrano et al 2018; Kefayati et al 2018) and the genetic dissection of important agronomical traits in walnut (Arab et al 2019; Famula et al 2019; Marrano et al 2019). However, the Chandler v1.0 assembly is highly fragmented, compromising the accuracy of gene prediction and the fulfillment of advanced genomics studies necessary to resolve many, still unanswered questions in walnut research.…”