Genotype-Corrector: improved genotype calls for genetic mapping in F2 and RIL populations

Effect of genotyping errors on linkage map construction based on repeated chip analysis of two recombinant inbred line populations in wheat (Triticum aestivum L.)

Linkage maps are essential for genetic mapping of phenotypic traits, gene map-based cloning, and marker-assisted selection in breeding applications. Construction of a high-quality saturated map requires high-quality genotypic data on a large number of molecular markers. Errors in genotyping cannot be completely avoided, no matter what platform is used. When genotyping error reaches a threshold level, it will seriously affect the accuracy of the constructed map and the reliability of consequent genetic studies. In this study, repeated genotyping of two recombinant inbred line (RIL) populations derived from crosses Yangxiaomai × Zhongyou 9507 and Jingshuang 16 × Bainong 64 was used to investigate the effect of genotyping errors on linkage map construction. Inconsistent data points between the two replications were regarded as genotyping errors, which were classified into three types. Genotyping errors were treated as missing values, and therefore the non-erroneous data set was generated. Firstly, linkage maps were constructed using the two replicates as well as the non-erroneous data set. Secondly, error correction methods implemented in software packages QTL IciMapping (EC) and Genotype-Corrector (GC) were applied to the two replicates. Linkage maps were therefore constructed based on the corrected genotypes and then compared with those from the non-erroneous data set. Simulation study was performed by considering different levels of genotyping errors to investigate the impact of errors and the accuracy of error correction methods. Results indicated that map length and marker order differed among the two replicates and the non-erroneous data sets in both RIL populations. For both actual and simulated populations, map length was expanded as the increase in error rate, and the correlation coefficient between linkage and physical maps became lower. Map quality can be improved by repeated genotyping and error correction algorithm. When it is impossible to genotype the whole mapping population repeatedly, 30% would be recommended in repeated genotyping. The EC method had a much lower false positive rate than did the GC method under different error rates. This study systematically expounded the impact of genotyping errors on linkage analysis, providing potential guidelines for improving the accuracy of linkage maps in the presence of genotyping errors.

Dissection of the E8 locus in two early maturing Canadian soybean populations

Soybean [Glycine max (L.) Merr.] is a short-day crop for which breeders want to expand the cultivation range to more northern agro-environments by introgressing alleles involved in early reproductive traits. To do so, we investigated quantitative trait loci (QTL) and expression quantitative trait loci (eQTL) regions comprised within the E8 locus, a large undeciphered region (~7.0 Mbp to 44.5 Mbp) associated with early maturity located on chromosome GM04. We used a combination of two mapping algorithms, (i) inclusive composite interval mapping (ICIM) and (ii) genome-wide composite interval mapping (GCIM), to identify major and minor regions in two soybean populations (QS15524F2:F3 and QS15544RIL) having fixed E1, E2, E3, and E4 alleles. Using this approach, we identified three main QTL regions with high logarithm of the odds (LODs), phenotypic variation explained (PVE), and additive effects for maturity and pod-filling within the E8 region: GM04:16,974,874-17,152,230 (E8-r1); GM04:35,168,111-37,664,017 (E8-r2); and GM04:41,808,599-42,376,237 (E8-r3). Using a five-step variant analysis pipeline, we identified Protein far-red elongated hypocotyl 3 (Glyma.04G124300; E8-r1), E1-like-a (Glyma.04G156400; E8-r2), Light-harvesting chlorophyll-protein complex I subunit A4 (Glyma.04G167900; E8-r3), and Cycling dof factor 3 (Glyma.04G168300; E8-r3) as the most promising candidate genes for these regions. A combinatorial eQTL mapping approach identified significant regulatory interactions for 13 expression traits (e-traits), including Glyma.04G050200 (Early flowering 3/E6 locus), with the E8-r3 region. Four other important QTL regions close to or encompassing major flowering genes were also detected on chromosomes GM07, GM08, and GM16. In GM07:5,256,305-5,404,971, a missense polymorphism was detected in the candidate gene Glyma.07G058200 (Protein suppressor of PHYA-105). These findings demonstrate that the locus known as E8 is regulated by at least three distinct genomic regions, all of which comprise major flowering genes.

Genetic analysis and QTL mapping of domestication-related traits in chili pepper (Capsicum annuum L.)

Chili pepper (Capsicum annuum L.) is one of the oldest and most phenotypically diverse pre-Columbian crops of the Americas. Despite the abundance of genetic resources, the use of wild germplasm and landraces in chili pepper breeding is limited. A better understanding of the evolutionary history in chili peppers, particularly in the context of traits of agronomic interest, can contribute to future improvement and conservation of genetic resources. In this study, an F₂:₃ mapping population derived from a cross between a C. annuum wild accession (Chiltepin) and a cultivated variety (Puya) was used to identify genomic regions associated with 19 domestication and agronomic traits. A genetic map was constructed consisting of 1023 single nucleotide polymorphism (SNP) markers clustered into 12 linkage groups and spanning a total of 1,263.87 cM. A reciprocal translocation that differentiates the domesticated genome from its wild ancestor and other related species was identified between chromosomes 1 and 8. Quantitative trait locus (QTL) analysis detected 20 marker-trait associations for 13 phenotypes, from which 14 corresponded to previously identified loci, and six were novel genomic regions related to previously unexplored domestication-syndrome traits, including form of unripe fruit, seedlessness, deciduous fruit, and growth habit. Our results revealed that the genetic architecture of Capsicum domestication is similar to other domesticated species with few loci with large effects, the presence of QTLs clusters in different genomic regions, and the predominance of domesticated recessive alleles. Our analysis indicates the domestication process in chili pepper has also had an effect on traits not directly related to the domestication syndrome. The information obtained in this study provides a more complete understanding of the genetic basis of Capsicum domestication that can potentially guide strategies for the exploitation of wild alleles.

Karyotype variation, spontaneous genome rearrangements affecting chemical insensitivity, and expression level polymorphisms in the plant pathogen Phytophthora infestans revealed using its first chromosome-scale assembly

Natural isolates of the potato and tomato pathogen Phytophthora infestans exhibit substantial variation in virulence, chemical sensitivity, ploidy, and other traits. A chromosome-scale assembly was developed to expand genomic resources for this oomyceteous microbe, and used to explore the basis of variation. Using PacBio and Illumina data, a long-range linking library, and an optical map, an assembly was created and coalesced into 15 pseudochromosomes spanning 219 Mb using SNP-based genetic linkage data. De novo gene prediction combined with transcript evidence identified 19,981 protein-coding genes, plus about eight thousand tRNA genes. The chromosomes were comprised of a mosaic of gene-rich and gene-sparse regions plus very long centromeres. Genes exhibited a biased distribution across chromosomes, especially members of families encoding RXLR and CRN effectors which clustered on certain chromosomes. Strikingly, half of F1 progeny of diploid parents were polyploid or aneuploid. Substantial expression level polymorphisms between strains were identified, much of which could be attributed to differences in chromosome dosage, transposable element insertions, and adjacency to repetitive DNA. QTL analysis identified a locus on the right arm of chromosome 3 governing sensitivity to the crop protection chemical metalaxyl. Strains heterozygous for resistance often experienced megabase-sized deletions of that part of the chromosome when cultured on metalaxyl, increasing resistance due to loss of the sensitive allele. This study sheds light on diverse phenomena affecting variation in P. infestans and relatives, helps explain the prevalence of polyploidy in natural populations, and provides a new foundation for biologic and genetic investigations.

The construction of a high-density consensus genetic map for soybean based on SNP markers derived from genotyping-by-sequencing

Genetic linkage maps are used to localize markers on the genome based on the recombination frequency. Most often, these maps are based on the segregation observed within a single biparental population of limited size (n < 300) where relatively few recombination events are sampled and in which some genomic regions are monomorphic because both parents carry the same alleles. Together, these two limitations affect both the resolution and extent of genome coverage of such maps. Consensus genetic maps overcome the limitations of individual genetic maps by merging the information from multiple segregating populations derived from a greater diversity of parental combinations, thus increasing the number of recombination events and reducing the number of monomorphic regions. The aim of this study was to construct a high-density consensus genetic map for single nucleotide polymorphism (SNP) markers obtained through a genotyping-by-sequencing (GBS) approach. Individual genetic maps were generated from six F4:5 mapping populations (n = 278-365), totaling 1857 individuals. The six linkage maps were then merged to produce a consensus map comprising a total of 16 311 mapped SNPs that jointly cover 99.5% of the soybean genome with only two gaps larger than 10 cM. Compared to previous soybean consensus maps, it offers a more extensive and uniform coverage.

A B73×Palomero Toluqueño mapping population reveals local adaptation in Mexican highland maize

Generations of farmer selection in the central Mexican highlands have produced unique maize varieties adapted to the challenges of the local environment. In addition to possessing great agronomic and cultural value, Mexican highland maize represents a good system for the study of local adaptation and acquisition of adaptive phenotypes under cultivation. In this study, we characterize a recombinant inbred line population derived from the B73 reference line and the Mexican highland maize variety Palomero Toluqueño. B73 and Palomero Toluqueño showed classic rank-changing differences in performance between lowland and highland field sites, indicative of local adaptation. Quantitative trait mapping identified genomic regions linked to effects on yield components that were conditionally expressed depending on the environment. For the principal genomic regions associated with ear weight and total kernel number, the Palomero Toluqueño allele conferred an advantage specifically in the highland site, consistent with local adaptation. We identified Palomero Toluqueño alleles associated with expression of characteristic highland traits, including reduced tassel branching, increased sheath pigmentation and the presence of sheath macrohairs. The oligogenic architecture of these three morphological traits supports their role in adaptation, suggesting they have arisen from consistent directional selection acting at distinct points across the genome. We discuss these results in the context of the origin of phenotypic novelty during selection, commenting on the role of de novo mutation and the acquisition of adaptive variation by gene flow from endemic wild relatives.

A high-resolution consensus linkage map for barley based on GBS-derived genotypes

As genotyping-by-sequencing (GBS) is widely used in barley genetic studies, the translation of the physical position of GBS-derived SNPs into accurate genetic positions has become relevant. The main aim of this study was to develop a high-resolution consensus linkage map based on GBS-derived SNPs. The construction of this integrated map involved 11 bi-parental populations composed of 3743 segregating progenies. We adopted a uniform set of SNP-calling and filtering conditions to identify 50 875 distinct SNPs segregating in at least one population. These SNPs were grouped into 18 580 non-redundant SNPs (bins). The resulting consensus linkage map spanned 1050.1 cM, providing an average density of 17.7 bins and 48.4 SNPs per cM. The consensus map is characterized by the absence of large intervals devoid of marker coverage (significant gaps), the largest interval between bins was only 3.7 cM and the mean distance between adjacent bins was 0.06 cM. This high-resolution linkage map will contribute to several applications in genomic research, such as providing useful information on the recombination landscape for QTLs/genes identified via GWAS or ensuring a uniform distribution of SNPs when developing low-cost genotyping tools offering a limited number of markers.

Novel QTL identification and candidate gene analysis for enhancing salt tolerance in soybean (Glycine max (L.) Merr.)

Soybean, a glycophyte that is sensitive to salt stress, is greatly affected by salinity at all growth stages. A mapping population derived from a cross between a salt-sensitive Korean cultivar, Cheongja 3, and a salt-tolerant landrace, IT162669, was used to identify quantitative trait loci (QTLs) conferring salt tolerance in soybean. Following treatment with 120 mM NaCl for 2 weeks, phenotypic traits representing physiological damage, leaf Na+ content, and K+/Na+ ratio were characterized. Among the QTLs mapped on a high-density genetic map harboring 2,630 single nucleotide polymorphism markers, we found two novel major loci, qST6, on chromosome 6, and qST10, on chromosome 10, which controlled traits related to ion toxicity and physiology in response to salinity, respectively. These loci were distinct from the previously known salt tolerance allele on chromosome 3. Other QTLs associated with abiotic stress overlapped with the genomic regions of qST6 and qST10, or with their paralogous regions. Based on the functional annotation and parental expression differences, we identified eight putative candidate genes, two in qST6 and six in qST10, which included a phosphoenolpyruvate carboxylase and an ethylene response factor. This study provides additional genetic resources to breed soybean cultivars with enhanced salt tolerance.

Identification of Candidate Genes Regulating the Seed Coat Color Trait in Sesame (Sesamum indicum L.) Using an Integrated Approach of QTL Mapping and Transcriptome Analysis

Seed coat color is an important seed quality trait in sesame. However, the genetic mechanism of seed coat color variation remains elusive in sesame. We conducted a QTL mapping of the seed coat color trait in sesame using an F₂ mapping population. With the aid of the newly constructed superdense genetic linkage map comprised of 22,375 bins distributed in 13 linkage groups (LGs), 17 QTLs of the three indices (i.e., L, a, and b values) of seed coat color were detected in seven intervals on four LGs, with a phenotype variance explanation rate of 4.46-41.53%. A new QTL qSCa6.1 on LG 6 and a QTL hotspot containing at least four QTLs on LG 9 were further identified. Variants screening of the target intervals showed that there were 84 genes which possessed the variants that were high-impact and co-segregating with the seed coat color trait. Meanwhile, we performed the transcriptome comparison of the developing seeds of a white- and a black-seeded variety, and found that the differentially expressed genes were significantly enriched in 37 pathways, including three pigment biosynthesis related pathways. Integration of variants screening and transcriptome comparison results suggested that 28 candidate genes probably participated in the regulation of the seed coat color in sesame; of which, 10 genes had been proved or suggested to be involved in pigments biosynthesis or accumulation during seed formation. The findings gave the basis for the mechanism of seed coat color regulation in sesame, and exhibited the effects of the integrated approach of genome resequencing and transcriptome analysis on the genetics analysis of the complex traits.

Genetic characterization of a Sorghum bicolor multiparent mapping population emphasizing carbon-partitioning dynamics

Sorghum bicolor, a photosynthetically efficient C₄ grass, represents an important source of grain, forage, fermentable sugars, and cellulosic fibers that can be utilized in myriad applications ranging from bioenergy to bioindustrial feedstocks. Sorghum’s efficient fixation of carbon per unit time per unit area per unit input has led to its classification as a preferred biomass crop highlighted by its designation as an advanced biofuel by the U.S. Department of Energy. Due to its extensive genetic diversity and worldwide colonization, sorghum has considerable diversity for a range of phenotypes influencing productivity, composition, and sink/source dynamics. To dissect the genetic basis of these key traits, we present a sorghum carbon-partitioning nested association mapping (NAM) population generated by crossing 11 diverse founder lines with Grassl as the single recurrent female. By exploiting existing variation among cellulosic, forage, sweet, and grain sorghum carbon partitioning regimes, the sorghum carbon-partitioning NAM population will allow the identification of important biomass-associated traits, elucidate the genetic architecture underlying carbon partitioning and improve our understanding of the genetic determinants affecting unique phenotypes within Poaceae. We contrast this NAM population with an existing grain population generated using Tx430 as the recurrent female. Genotypic data are assessed for quality by examining variant density, nucleotide diversity, linkage decay, and are validated using pericarp and testa phenotypes to map known genes affecting these phenotypes. We release the 11-family NAM population along with corresponding genomic data for use in genetic, genomic, and agronomic studies with a focus on carbon-partitioning regimes.

Discovery of a Major QTL Controlling Trichome IV Density in Tomato Using K-Seq Genotyping

Trichomes are a common morphological defense against pests, in particular, type IV glandular trichomes have been associated with resistance against different invertebrates. Cultivated tomatoes usually lack or have a very low density of type IV trichomes. Therefore, for sustainable management of this crop, breeding programs could incorporate some natural defense mechanisms, such as those afforded by trichomes, present in certain Solanum species. We have identified a S. pimpinellifolium accession with very high density of this type of trichomes. This accession was crossed with a S. lycopersicum var. cerasiforme and a S. lycopersicum var. lycopersicum accessions, and the two resulting F2 populations have been characterized and genotyped using a new genotyping methodology, K-seq. We have been able to build an ultra-dense genetic map with 147,326 SNP markers with an average distance between markers of 0.2 cm that has allowed us to perform a detailed mapping. We have used two different families and two different approaches, QTL mapping and QTL-seq, to identify several QTLs implicated in the control of trichome type IV developed in this accession on the chromosomes 5, 6, 9 and 11. The QTL located on chromosome 9 is a major QTL that has not been previously reported in S. pimpinellifolium. This QTL could be easily introgressed in cultivated tomato due to the close genetic relationship between both species.

Optimizing imputation of marker data from genotyping-by-sequencing (GBS) for genomic selection in non-model species: Rubber tree (Hevea brasiliensis) as a case study

Genotyping-by-sequencing (GBS) provides the marker density required for genomic predictions (GP). However, GBS gives a high proportion of missing SNP data which, for species without a chromosome-level genome assembly, must be imputed without knowing the SNP physical positions. Here, we compared GP accuracy with seven map-independent and two map-dependent imputation approaches, and when using all SNPs against the subset of genetically mapped SNPs. We used two rubber tree (Hevea brasiliensis) datasets with three traits. The results showed that the best imputation approaches were LinkImputeR, Beagle and FImpute. Using the genetically mapped SNPs increased GP accuracy by 4.3%. Using LinkImputeR on all the markers allowed avoiding genetic mapping, with a slight decrease in GP accuracy. LinkImputeR gave the highest level of correctly imputed genotypes and its performances were further improved by its ability to define a subset of SNPs imputed optimally. These results will contribute to the efficient implementation of genomic selection with GBS. For Hevea, GBS is promising for rubber yield improvement, with GP accuracies reaching 0.52.

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.

QTL mapping of spontaneous haploid genome doubling using genotyping-by-sequencing in maize (Zea mays L.)

A major QTL for SHGD was identified on chromosome 5 with stable expression across environments. The introgression this QTL can overcome the need of colchicine in DH lines development. Genome doubling of haploids is one of the major constraints of large-scale doubled haploid (DH) technology. Improving spontaneous haploid genome doubling (SHGD) is an alternative to overcome this limitation. In this study, we aimed to construct a high-density linkage map based on genotyping by sequencing of single nucleotide polymorphism, to detect QTL and QTL by environment (Q by E) interactions affecting SHGD and to identify the best trait for mapping and selection of haploid male fertility (HMF). To this end, a biparental population of 220 F_2:3 families was developed from a cross between A427 (high HMF) and CR1Ht (moderate HMF) to be used as donor. A high-density linkage map was constructed containing 4171 SNP markers distributed over 10 chromosomes with an average distance between adjacent markers of 0.51 cM. QTL mapping for haploid fertile anther emergence, pollen production, tassel size, and HMF, identified 27 QTL across three environments, and Q by E interactions were significant. A major QTL was identified on chromosome 5. This QTL explained over 45% of the observed variance for all traits across all environments. The introgression of this major QTL, using marker-assisted backcrossing, has great potential to overcome the need of using colchicine in DH line development.

IonBreeders: bioinformatics plugins toward genomics-assisted breeding

Polymorphism information generated by next-generation sequencing (NGS) technologies has enabled applications of genome-wide markers assisted breeding. However, handling such large-scale data remains a challenge for experimental researchers and breeders, calling for the urgent development of a flexible and straightforward analysis tool for NGS data. We developed "IonBreeders" as bioinformatics plugins that implement general analysis steps from genotyping to genomic prediction. IonBreeders comprises three plugins, "ABH", "IMPUTATION", and "GENOMIC PREDICTION", for format conversion of genotyping data, preprocessing and imputation of genotyping data, and genomic prediction, respectively. "ABH" converts genotyping data derived from NGS into the ABH format, which is acceptable for our further plugins and with other breeding software tools, R/qtl, MapMaker, and AntMap. "IMPUTATION" filters out non-informative markers and imputes missing marker genotypes. In "GENOMIC PREDICTION", users can use four statistical methods based on their target trait, quantitative trait locus effect, and number of markers, and construct a prediction model for genomic selection. IonBreeders is operated in Torrent Suite, but can also handle genotype data in standard formats, e.g., Variant Call Format (VCF), by format conversion using free software or our provided scripts.

Intraspecific Genetic Variation Underlying Postmating Reproductive Barriers between Species in the Wild Tomato Clade (Solanum sect. Lycopersicon)

A goal of speciation genetics is to understand how the genetic components underlying interspecific reproductive barriers originate within species. Unilateral incompatibility (UI) is a postmating prezygotic barrier in which pollen rejection in the female reproductive tract (style) occurs in only one direction of an interspecific cross. Natural variation in the strength of UI has been observed among populations within species in the wild tomato clade. In some cases, molecular loci underlying self-incompatibility (SI) are associated with this variation in UI, but the mechanistic connection between these intra- and inter-specific pollen rejection behaviors is poorly understood in most instances. We generated an F2 population between SI and SC genotypes of a single species, Solanum pennellii, to examine the genetic basis of intraspecific variation in UI against other species, and to determine whether loci underlying SI are genetically associated with this variation. We found that F2 individuals vary in the rate at which UI rejection occurs. One large effect QTL detected for this trait co-localized with the SI-determining S-locus. Moreover, individuals that expressed S-RNase-the S-locus protein involved in SI pollen rejection-in their styles had much more rapid UI responses compared with those without S-RNase protein. Our analysis shows that intraspecific variation at mate choice loci-in this case at loci that prevent self-fertilization-can contribute to variation in the expression of interspecific isolation, including postmating prezygotic barriers. Understanding the nature of such intraspecific variation can provide insight into the accumulation of these barriers between diverging lineages.

Intraspecific Genetic Variation Underlying Postmating Reproductive Barriers between Species in the Wild Tomato Clade (Solanum sect. Lycopersicon)

A goal of speciation genetics is to understand how the genetic components underlying interspecific reproductive barriers originate within species. Unilateral incompatibility (UI) is a postmating prezygotic barrier in which pollen rejection in the female reproductive tract (style) occurs in only one direction of an interspecific cross. Natural variation in the strength of UI has been observed among populations within species in the wild tomato clade. In some cases, molecular loci underlying self-incompatibility (SI) are associated with this variation in UI, but the mechanistic connection between these intra- and inter-specific pollen rejection behaviors is poorly understood in most instances. We generated an F2 population between SI and SC genotypes of a single species, Solanum pennellii, to examine the genetic basis of intraspecific variation in UI against other species, and to determine whether loci underlying SI are genetically associated with this variation. We found that F2 individuals vary in the rate at which UI rejection occurs. One large effect QTL detected for this trait co-localized with the SI-determining S-locus. Moreover, individuals that expressed S-RNase-the S-locus protein involved in SI pollen rejection-in their styles had much more rapid UI responses compared with those without S-RNase protein. Our analysis shows that intraspecific variation at mate choice loci-in this case at loci that prevent self-fertilization-can contribute to variation in the expression of interspecific isolation, including postmating prezygotic barriers. Understanding the nature of such intraspecific variation can provide insight into the accumulation of these barriers between diverging lineages.