Evaluation and Recommendations for Routine Genotyping Using Skim Whole Genome Re-sequencing in Canola

A multitiered haplotype strategy to enhance phased assembly and fine mapping of a disease resistance locus

Fine mapping of quantitative trait loci (QTL) to dissect the genetic basis of traits of interest is essential to modern breeding practice. Here, we employed a multitiered haplotypic marker system to increase fine mapping accuracy by constructing a chromosome-level, haplotype-resolved parental genome, accurate detection of recombination sites, and allele-specific characterization of the transcriptome. In the first tier of this system, we applied the preexisting panel of 2,000 rhAmpSeq core genome markers that is transferable across the entire Vitis genus and provides a genomic resolution of 200 kb to 1 Mb. The second tier consisted of high-density haplotypic markers generated from Illumina skim sequencing data for samples enriched for relevant recombinations, increasing the potential resolution to hundreds of base pairs. We used this approach to dissect a novel Resistance to Plasmopara viticola-33 (RPV33) locus conferring resistance to grapevine downy mildew, narrowing the candidate region to only 0.46 Mb. In the third tier, we used allele-specific RNA-seq analysis to identify a cluster of 3 putative disease resistance RPP13-like protein 2 genes located tandemly in a nonsyntenic insertion as candidates for the disease resistance trait. In addition, combining the rhAmpSeq core genome haplotype markers and skim sequencing-derived high-density haplotype markers enabled chromosomal-level scaffolding and phasing of the grape Vitis × doaniana 'PI 588149' assembly, initially built solely from Pacific Biosciences (PacBio) high-fidelity (HiFi) reads, leading to the correction of 16 large-scale phasing errors. Our mapping strategy integrates high-density, phased genetic information with individual reference genomes to pinpoint the genetic basis of QTLs and will likely be widely adopted in highly heterozygous species.

Genetic Analysis and Physical Mapping of Oat Adult Plant Resistance Loci Against Puccinia coronata f. sp. avenae

Six quantitative trait loci (QTLs) for adult plant resistance against oat crown rust (Puccinia coronata f. sp. avenae) were identified from mapping three recombinant inbred populations. Using genotyping-by-sequencing with markers called against the OT3098 v1 reference genome, the QTLs were mapped on six different chromosomes: Chr1D, Chr4D, Chr5A, Chr5D, Chr7A, and Chr7C. Composite interval mapping with marker cofactor selection showed that the phenotypic variance explained by all identified QTLs for coefficient of infection range from 12.2 to 46.9%, whereas heritability estimates ranged from 0.11 to 0.38. The significant regions were narrowed down to intervals of 3.9 to 25 cM, equivalent to physical distances of 11 to 133 Mb. At least two flanking single-nucleotide polymorphism markers were identified within 10 cM of each QTL that could be used in marker-assisted introgression, pyramiding, and selection. The additive effects of the QTLs in each population were determined using single-nucleotide polymorphism haplotype data, which showed a significantly lower coefficient of infection in lines homozygous for the resistant alleles. Analysis of pairwise linkage disequilibrium also revealed high correlation of markers and presence of linkage blocks in the significant regions. To further facilitate marker-assisted breeding, polymerase chain reaction allelic competitive extension (PACE) markers for the adult plant resistance loci were developed. Putative candidate genes were also identified in each of the significant regions, which include resistance gene analogs that encode for kinases, ligases, and predicted receptors of avirulence proteins from pathogens.

Skim resequencing finely maps the downy mildew resistance loci RPF2 and RPF3 in spinach cultivars whale and Lazio

Commercial production of spinach (Spinacia oleracea L.) is centered in California and Arizona in the US, where downy mildew caused by Peronospora effusa is the most destructive disease. Nineteen typical races of P. effusa have been reported to infect spinach, with 16 identified after 1990. The regular appearance of new pathogen races breaks the resistance gene introgressed in spinach. We attempted to map and delineate the RPF2 locus at a finer resolution, identify linked single nucleotide polymorphism (SNP) markers, and report candidate downy mildew resistance (R) genes. Progeny populations segregating for RPF2 locus derived from resistant differential cultivar Lazio were infected using race 5 of P. effusa and were used to study for genetic transmission and mapping analysis in this study. Association analysis performed with low coverage whole genome resequencing-generated SNP markers mapped the RPF2 locus between 0.47 to 1.46 Mb of chromosome 3 with peak SNP (Chr3_1, 221, 009) showing a LOD value of 61.6 in the GLM model in TASSEL, which was within 1.08 Kb from Spo12821, a gene that encodes CC-NBS-LRR plant disease resistance protein. In addition, a combined analysis of progeny panels of Lazio and Whale segregating for RPF2 and RPF3 loci delineated the resistance section in chromosome 3 between 1.18-1.23 and 1.75-1.76 Mb. This study provides valuable information on the RPF2 resistance region in the spinach cultivar Lazio compared to RPF3 loci in the cultivar Whale. The RPF2 and RPF3 specific SNP markers, plus the resistant genes reported here, could add value to breeding efforts to develop downy mildew resistant cultivars in the future.

Discovery and genome-guided mapping of REN12 from Vitis amurensis, conferring strong, rapid resistance to grapevine powdery mildew

Powdery mildew resistance genes restrict infection attempts at different stages of pathogenesis. Here, a strong and rapid powdery mildew resistance phenotype was discovered from Vitis amurensis 'PI 588631' that rapidly stopped over 97% of Erysiphe necator conidia, before or immediately after emergence of a secondary hypha from appressoria. This resistance was effective across multiple years of vineyard evaluation on leaves, stems, rachises, and fruit and against a diverse array of E. necator laboratory isolates. Using core genome rhAmpSeq markers, resistance mapped to a single dominant locus (here named REN12) on chromosome 13 near 22.8-27.0 Mb, irrespective of tissue type, explaining up to 86.9% of the phenotypic variation observed on leaves. Shotgun sequencing of recombinant vines using skim-seq technology enabled the locus to be further resolved to a 780 kb region, from 25.15 to 25.93 Mb. RNASeq analysis indicated the allele-specific expression of four resistance genes (NLRs) from the resistant parent. REN12 is one of the strongest powdery mildew resistance loci in grapevine yet documented, and the rhAmpSeq sequences presented here can be directly used for marker-assisted selection or converted to other genotyping platforms. While no virulent isolates were identified among the genetically diverse isolates and wild populations of E. necator tested here, NLR loci like REN12 are often race-specific. Thus, stacking of multiple resistance genes and minimal use of fungicides should enhance the durability of resistance and could enable a 90% reduction in fungicides in low-rainfall climates where few other pathogens attack the foliage or fruit.

Innovative Advances in Plant Genotyping

Over the past decade, advances in plant genotyping have been critical in enabling the identification of genetic diversity, in understanding evolution, and in dissecting important traits in both crops and native plants. The widespread popularity of single-nucleotide polymorphisms (SNPs) has prompted significant improvements to SNP-based genotyping, including SNP arrays, genotyping by sequencing, and whole-genome resequencing. More recent approaches, including genotyping structural variants, utilizing pangenomes to capture species-wide genetic diversity and exploiting machine learning to analyze genotypic data sets, are pushing the boundaries of what plant genotyping can offer. In this chapter, we highlight these innovations and discuss how they will accelerate and advance future genotyping efforts.

Genomic architecture of migration timing in a long-distance migratory songbird

The impact of climate change on spring phenology poses risks to migratory birds, as migration timing is controlled predominantly by endogenous mechanisms. Despite recent advances in our understanding of the underlying genetic basis of migration timing, the ways that migration timing phenotypes in wild individuals may map to specific genomic regions requires further investigation. We examined the genetic architecture of migration timing in a long-distance migratory songbird (purple martin, Progne subis subis) by integrating genomic data with an extensive dataset of direct migratory tracks. A moderate to large amount of variance in spring migration arrival timing was explained by genomics (proportion of phenotypic variation explained by genomics = 0.74; polygenic score R² = 0.24). On chromosome 1, a region that was differentiated between migration timing phenotypes contained genes that could facilitate nocturnal flights and act as epigenetic modifiers. Overall, these results advance our understanding of the genomic underpinnings of migration timing.

Construction of ultra-high-density genetic linkage map of a sorghum-sudangrass hybrid using whole genome resequencing

The sorghum-sudangrass hybrid is a vital annual gramineous herbage. Few reports exist on its ultra-high-density genetic map. In this study, we sought to create an ultra-high-density genetic linkage map for this hybrid to strengthen its functional genomics research and genetic breeding. We used 150 sorghum-sudangrass hybrid F2 individuals and their parents (scattered ear sorghum and red hull sudangrass) for high-throughput sequencing on the basis of whole genome resequencing. In total, 1,180.66 Gb of data were collected. After identification, filtration for integrity, and partial segregation, over 5,656 single nucleotide polymorphism markers of high quality were detected. An ultra-high-density genetic linkage map was constructed using these data. The markers covered approximately 2,192.84 cM of the map with average marker intervals of 0.39 cM. The length ranged from 115.39 cM to 264.04 cM for the 10 linkage groups. Currently, this represents the first genetic linkage map of this size, number of molecular markers, density, and coverage for sorghum-sudangrass hybrid. The findings of this study provide valuable genome-level information on species evolution and comparative genomics analysis and lay the foundation for further research on quantitative trait loci fine mapping and gene cloning and marker-assisted breeding of important traits in sorghum-sudangrass hybrids.

Mapping and selection of downy mildew resistance in spinach cv. whale by low coverage whole genome sequencing

Spinach (Spinacia oleracea) is a popular leafy vegetable crop and commercial production is centered in California and Arizona in the US. The oomycete Peronospora effusa causes the most important disease in spinach, downy mildew. A total of nineteen races of P. effusa are known, with more than 15 documented in the last three decades, and the regular emergence of new races is continually overcoming the genetic resistance to the pathogen. This study aimed to finely map the downy mildew resistance locus RPF3 in spinach, identify single nucleotide polymorphism (SNP) markers associated with the resistance, refine the candidate genes responsible for the resistance, and evaluate the prediction performance using multiple machine learning genomic prediction (GP) methods. Segregating progeny population developed from a cross of resistant cultivar Whale and susceptible cultivar Viroflay to race 5 of P. effusa was inoculated under greenhouse conditions to determine downy mildew disease response across the panel. The progeny panel and the parents were resequenced at low coverage (1x) to identify genome wide SNP markers. Association analysis was performed using disease response phenotype data and SNP markers in TASSEL, GAPIT, and GENESIS programs and mapped the race 5 resistance loci (RPF3) to 1.25 and 2.73 Mb of Monoe-Viroflay chromosome 3 with the associated SNP in the 1.25 Mb region was 0.9 Kb from the NBS-LRR gene SOV3g001250. The RPF3 locus in the 1.22-1.23 Mb region of Sp75 chromosome 3 is 2.41-3.65 Kb from the gene Spo12821 annotated as NBS-LRR disease resistance protein. This study extended our understanding of the genetic basis of downy mildew resistance in spinach cultivar Whale and mapped the RPF3 resistance loci close to the NBS-LRR gene providing a target to pursue functional validation. Three SNP markers efficiently selected resistance based on multiple genomic selection (GS) models. The results from this study have added new genomic resources, generated an informed basis of the RPF3 locus resistant to spinach downy mildew pathogen, and developed markers and prediction methods to select resistant lines.

Genomic regions associated with virulence in Setosphaeria turcica identified by linkage mapping in a biparental population

Northern corn leaf blight (NCLB) and sorghum leaf blight (SLB) are significant diseases of maize and sorghum, respectively, caused by the filamentous fungus Setosphaeria turcica. Strains of S. turcica are typically host-specific and infect either maize or sorghum. Host specificity in this pathogen is attributed to a single locus for maize and a second distinct locus for sorghum. To identify the genetic basis of host specificity in S. turcica, we generated a biparental population of S. turcica by crossing strains specific to maize and sorghum, phenotyped the population for leaf blight on sorghum and maize, genotyped the population to create a linkage map of S. turcica, and located candidate virulence regions. A total of 190 ascospores from 35 pseudothecia were isolated from the cross of maize and sorghum-specific strains. Greenhouse phenotyping of the biparental population (n = 144) showed independent inheritance of virulence, as indicated by a 1:1:1:1 segregation for virulence to maize, sorghum, both maize and sorghum, and avirulence to both crops. The population and host-specific parent strains were genotyped using genome skim sequencing on an Illumina NovaSeq 6000 platform resulting in over 780 million reads. A total of 32,635 variants including single nucleotide polymorphisms and indels were scored. There was evidence for a large deletion in the sorghum-specific strain of S. turcica. A genetic map consisting of 17 linkage groups spanning 3,069 centimorgans was constructed. Virulence to sorghum and maize mapped on distinct linkage groups with a significant QTL detected for virulence to maize. Furthermore, a single locus each for the in vitro traits hyphal growth rate and conidiation were identified and mapped onto two other linkage groups. In vitro traits did not correlate with in planta virulence complexity, suggesting that virulence on both hosts does not incur a fitness cost. Hyphal growth rate and conidiation were negatively correlated, indicating differences in hyphal growth versus dispersal ability for this pathogen. Identification of genetic regions underlying virulence specificity and saprotrophic growth traits in S. turcica provides a better understanding of the S. turcica- Andropogoneae pathosystem.

Development of the Wheat Practical Haplotype Graph Database as a Resource for Genotyping Data Storage and Genotype Imputation

To improve the efficiency of high-density genotype data storage and imputation in bread wheat (Triticum aestivum L.), we applied the Practical Haplotype Graph (PHG) tool. The Wheat PHG database was built using whole-exome capture sequencing data from a diverse set of 65 wheat accessions. Population haplotypes were inferred for the reference genome intervals defined by the boundaries of the high-quality gene models. Missing genotypes in the inference panels, composed of wheat cultivars or recombinant inbred lines genotyped by exome capture, genotyping-by-sequencing (GBS), or whole-genome skim-seq sequencing approaches, were imputed using the Wheat PHG database. Though imputation accuracy varied depending on the method of sequencing and coverage depth, we found 92% imputation accuracy with 0.01× sequence coverage, which was slightly lower than the accuracy obtained using the 0.5× sequence coverage (96.6%). Compared to Beagle, on average, PHG imputation was ∼3.5% (P-value < 2 × 10⁻¹⁴) more accurate, and showed 27% higher accuracy at imputing a rare haplotype introgressed from a wild relative into wheat. We found reduced accuracy of imputation with independent 2× GBS data (88.6%), which increases to 89.2% with the inclusion of parental haplotypes in the database. The accuracy reduction with GBS is likely associated with the small overlap between GBS markers and the exome capture dataset, which was used for constructing PHG. The highest imputation accuracy was obtained with exome capture for the wheat D genome, which also showed the highest levels of linkage disequilibrium and proportion of identity-by-descent regions among accessions in the PHG database. We demonstrate that genetic mapping based on genotypes imputed using PHG identifies SNPs with a broader range of effect sizes that together explain a higher proportion of genetic variance for heading date and meiotic crossover rate compared to previous studies.

Impact of genotypic errors with equal and unequal family contribution on accuracy of genomic prediction in aquaculture using simulation

Genotypic errors, conflict between recorded genotype and the true genotype, can lead to false or biased population genetic parameters. Here, the effect of genotypic errors on accuracy of genomic predictions and genomic relationship matrix are investigated using a simulation study based on population and genomic structure comparable to black tiger prawn, Penaeus monodon. Fifty full-sib families across five generations with phenotypic and genotypic information on 53 K SNPs were simulated. Ten replicates of different scenarios with three heritability estimates, equal and unequal family contributions were generated. Within each scenario, four SNP densities and three genotypic error rates in each SNP density were implemented. Results showed that family contribution did not have a substantial impact on accuracy of predictions across different datasets. In the absence of genotypic errors, 3 K SNP density was found to be efficient in estimating the accuracy, whilst increasing the SNP density from 3 to 20 K resulted in a marginal increase in accuracy of genomic predictions using the current population and genomic parameters. In addition, results showed that the presence of even 10% errors in a 10 and 20 K SNP panel might not have a severe impact on accuracy of predictions. However, below 10 K marker density, even a 5% error can result in lower accuracy of predictions.

Linkage mapping and QTL analysis of flowering time using ddRAD sequencing with genotype error correction in Brassica napus

BACKGROUND

Brassica napus is an important oilseed crop cultivated worldwide. During domestication and breeding of B. napus, flowering time has been a target of selection because of its substantial impact on yield. Here we use double digest restriction-site associated DNA sequencing (ddRAD) to investigate the genetic basis of flowering in B. napus. An F2 mapping population was derived from a cross between an early-flowering spring type and a late-flowering winter type.

RESULTS

Flowering time in the mapping population differed by up to 25 days between individuals. High genotype error rates persisted after initial quality controls, as suggested by a genotype discordance of ~ 12% between biological sequencing replicates. After genotype error correction, a linkage map spanning 3981.31 cM and compromising 14,630 single nucleotide polymorphisms (SNPs) was constructed. A quantitative trait locus (QTL) on chromosome C2 was detected, covering eight flowering time genes including FLC.

CONCLUSIONS

These findings demonstrate the effectiveness of the ddRAD approach to sample the B. napus genome. Our results also suggest that ddRAD genotype error rates can be higher than expected in F2 populations. Quality filtering and genotype correction and imputation can substantially reduce these error rates and allow effective linkage mapping and QTL analysis.

Fast-GBS v2.0: an analysis toolkit for genotyping-by-sequencing data

Genotyping-by-sequencing (GBS) is a rapid, flexible, low-cost, and robust genotyping method that simultaneously discovers variants and calls genotypes within a broad range of samples. These characteristics make GBS an excellent tool for many applications and research questions from conservation biology to functional genomics in both model and non-model species. Continued improvement of GBS relies on a more comprehensive understanding of data analysis, development of fast and efficient bioinformatics pipelines, accurate missing data imputation, and active post-release support. Here, we present the second generation of Fast-GBS (v2.0) that offers several new options (e.g., processing paired-end reads and imputation of missing data) and features (e.g., summary statistics of genotypes) to improve the GBS data analysis process. The performance assessment analysis showed that Fast-GBS v2.0 outperformed other available analytical pipelines, such as GBS-SNP-CROP and Gb-eaSy. Fast-GBS v2.0 provides an analysis platform that can be run with different types of sequencing data, modest computational resources, and allows for missing-data imputation for various species in different contexts.

The Use of Genetic and Gene Technologies in Shaping Modern Rapeseed Cultivars (Brassica napus L.)

Since their domestication, Brassica oilseed species have undergone progressive transformation allied with the development of breeding and molecular technologies. The canola (Brassica napus) crop has rapidly expanded globally in the last 30 years with intensive innovations in canola varieties, providing for a wider range of markets apart from the food industry. The breeding efforts of B. napus, the main source of canola oil and canola meal, have been mainly focused on improving seed yield, oil quality, and meal quality along with disease resistance, abiotic stress tolerance, and herbicide resistance. The revolution in genetics and gene technologies, including genetic mapping, molecular markers, genomic tools, and gene technology, especially gene editing tools, has allowed an understanding of the complex genetic makeup and gene functions in the major bioprocesses of the Brassicales, especially Brassica oil crops. Here, we provide an overview on the contributions of these technologies in improving the major traits of B. napus and discuss their potential use to accomplish new improvement targets.

Meta-analysis of GWAS in canola blackleg (Leptosphaeria maculans) disease traits demonstrates increased power from imputed whole-genome sequence

Blackleg disease causes yield losses in canola (Brassica napus L.). To identify resistance genes and genomic regions, genome-wide association studies (GWAS) of 585 diverse winter and spring canola accessions were performed using imputed whole-genome sequence (WGS) and transcriptome genotype-by-sequencing (GBSt). Blackleg disease phenotypes were collected across three years in six trials. GWAS were performed in several ways and their respective power was judged by the number of significant single nucleotide polymorphisms (SNP), the false discovery rate (FDR), and the percentage of SNP that validated in additional field trials in two subsequent years. WGS GWAS with 1,234,708 million SNP detected a larger number of significant SNP, achieved a lower FDR and a higher validation rate than GBSt with 64,072 SNP. A meta-analysis combining survival and average internal infection resulted in lower FDR but also lower validation rates. The meta-analysis GWAS identified 79 genomic regions (674 SNP) conferring potential resistance to L. maculans. While several GWAS signals localised in regions of known Rlm genes, fifty-three new potential resistance regions were detected. Seventeen regions had underlying genes with putative functions related to disease defence or stress response in Arabidopsis thaliana. This study provides insight into the genetic architecture and potential molecular mechanisms underlying canola L. maculans resistance.

Evaluation of variant calling tools for large plant genome re-sequencing

BACKGROUND

Discovering single nucleotide polymorphisms (SNPs) from agriculture crop genome sequences has been a widely used strategy for developing genetic markers for several applications including marker-assisted breeding, population diversity studies for eco-geographical adaption, genotyping crop germplasm collections, and others. Accurately detecting SNPs from large polyploid crop genomes such as wheat is crucial and challenging. A few variant calling methods have been previously developed but they show a low concordance between their variant calls. A gold standard of variant sets generated from one human individual sample was established for variant calling tool evaluations, however hitherto no gold standard of crop variant set is available for wheat use. The intent of this study was to evaluate seven SNP variant calling tools (FreeBayes, GATK, Platypus, Samtools/mpileup, SNVer, VarScan, VarDict) with the two most popular mapping tools (BWA-mem and Bowtie2) on wheat whole exome capture (WEC) re-sequencing data from allohexaploid wheat.

RESULTS

We found the BWA-mem mapping tool had both a higher mapping rate and a higher accuracy rate than Bowtie2. With the same mapping quality (MQ) cutoff, BWA-mem detected more variant bases in mapping reads than Bowtie2. The reads preprocessed with quality trimming or duplicate removal did not significantly affect the final mapping performance in terms of mapped reads. Based on the concordance and receiver operating characteristic (ROC), the Samtools/mpileup variant calling tool with BWA-mem mapping of raw sequence reads outperformed other tests followed by FreeBayes and GATK in terms of specificity and sensitivity. VarDict and VarScan were the poorest performing variant calling tools with the wheat WEC sequence data.

CONCLUSION

The BWA-mem and Samtools/mpileup pipeline, with no need to preprocess the raw read data before mapping onto the reference genome, was ascertained the optimum for SNP calling for the complex wheat genome re-sequencing. These results also provide useful guidelines for reliable variant identification from deep sequencing of other large polyploid crop genomes.

Construction of a High-Density Genetic Map and Mapping of Firmness in Grapes (Vitis vinifera L.) Based on Whole-Genome Resequencing

Berry firmness is one of the most important quality traits in table grapes. The underlying molecular and genetic mechanisms for berry firmness remain unclear. We constructed a high-density genetic map based on whole-genome resequencing to identify loci associated with berry firmness. The genetic map had 19 linkage groups, including 1662 bin markers (26,039 SNPs), covering 1463.38 cM, and the average inter-marker distance was 0.88 cM. An analysis of berry firmness in the F1 population and both parents for three consecutive years revealed continuous variability in F1, with a distribution close to the normal distribution. Based on the genetic map and phenotypic data, three potentially significant quantitative trait loci (QTLs) related to berry firmness were identified by composite interval mapping. The contribution rate of each QTL ranged from 21.5% to 28.6%. We identified four candidate genes associated with grape firmness, which are related to endoglucanase, abscisic acid (ABA), and transcription factors. A qRT-PCR analysis revealed that the expression of abscisic-aldehyde oxidase-like gene (VIT_18s0041g02410) and endoglucanase 3 gene (VIT_18s0089g00210) in Muscat Hamburg was higher than in Crimson Seedless at the veraison stage, which was consistent with that of parent berry firmness. These results confirmed that VIT_18s0041g02410 and VIT_18s0089g00210 are candidate genes associated with berry firmness.

Assessment of low-coverage nanopore long read sequencing for SNP genotyping in doubled haploid canola (Brassica napus L.)

Despite the high accuracy of short read sequencing (SRS), there are still issues with attaining accurate single nucleotide polymorphism (SNP) genotypes at low sequencing coverage and in highly duplicated genomes due to misalignment. Long read sequencing (LRS) systems, including the Oxford Nanopore Technologies (ONT) minION, have become popular options for de novo genome assembly and structural variant characterisation. The current high error rate often requires substantial post-sequencing correction and would appear to prevent the adoption of this system for SNP genotyping, but nanopore sequencing errors are largely random. Using low coverage ONT minION sequencing for genotyping of pre-validated SNP loci was examined in 9 canola doubled haploids. The minION genotypes were compared to the Illumina sequences to determine the extent and nature of genotype discrepancies between the two systems. The significant increase in read length improved alignment to the genome and the absence of classical SRS biases results in a more even representation of the genome. Sequencing errors are present, primarily in the form of heterozygous genotypes, which can be removed in completely homozygous backgrounds but requires more advanced bioinformatics in heterozygous genomes. Developments in this technology are promising for routine genotyping in the future.