Clusthaplo: a plug-in for MCQTL to enhance QTL detection using ancestral alleles in multi-cross design

WGNAM: whole-genome nested association mapping

A powerful QTL analysis method for nested association mapping populations is presented. Based on a one-stage multi-locus model, it provides accurate predictions of founder specific QTL effects. Nested association mapping (NAM) populations have been created to enable the identification of quantitative trait loci (QTL) in different genetic backgrounds. A whole-genome nested association mapping (WGNAM) method is presented to perform QTL analysis in NAM populations. The WGNAM method is an adaptation of the multi-parent whole genome average interval mapping approach where the crossing design is incorporated through the probability of inheriting founder alleles for every marker across the genome. Based on a linear mixed model, this method provides a one-stage analysis of raw phenotypic data, molecular markers, and crossing design. It simultaneously scans the whole-genome through an iterative process leading to a model with all the identified QTL while keeping the false positive rate low. The WGNAM approach was assessed through a simulation study, confirming to be a powerful and accurate method for QTL analysis for a NAM population. This novel method can also accommodate a multi-reference NAM (MR-NAM) population where donor parents are crossed with multiple reference parents to increase genetic diversity. Therefore, a demonstration is presented using a MR-NAM population for wheat (Triticum aestivum L.) to perform a QTL analysis for plant height. The strength and size of the putative QTL were summarized enhancing the understanding of the QTL effects depending on the parental origin. Compared to other methods, the proposed methodology based on a one-stage analysis provides greater power to detect QTL and increased accuracy in the estimation of their effects. The WGNAM method establishes the basis for accurate QTL mapping studies for NAM and MR-NAM populations.

Modeling allelic diversity of multiparent mapping populations affects detection of quantitative trait loci

The search for quantitative trait loci that explain complex traits such as yield and drought tolerance has been ongoing in all crops. Methods such as biparental quantitative trait loci mapping and genome-wide association studies each have their own advantages and limitations. Multiparent advanced generation intercross populations contain more recombination events and genetic diversity than biparental mapping populations and are better able to estimate effect sizes of rare alleles than association mapping populations. Here, we discuss the results of using a multiparent advanced generation intercross population of doubled haploid maize lines created from 16 diverse founders to perform quantitative trait loci mapping. We compare 3 models that assume bi-allelic, founder, and ancestral haplotype allelic states for quantitative trait loci. The 3 methods have differing power to detect quantitative trait loci for a variety of agronomic traits. Although the founder approach finds the most quantitative trait loci, all methods are able to find unique quantitative trait loci, suggesting that each model has advantages for traits with different genetic architectures. A closer look at a well-characterized flowering time quantitative trait loci, qDTA8, which contains vgt1, highlights the strengths and weaknesses of each method and suggests a potential epistatic interaction. Overall, our results reinforce the importance of considering different approaches to analyzing genotypic datasets, and shows the limitations of binary SNP data for identifying multiallelic quantitative trait loci.

Multiallelic models for QTL mapping in diverse polyploid populations

Abstract

Quantitative trait locus (QTL) analysis allows to identify regions responsible for a trait and to associate alleles with their effect on phenotypes. When using biallelic markers to find these QTL regions, two alleles per QTL are modelled. This assumption might be close to reality in specific biparental crosses but is unrealistic in situations where broader genetic diversity is studied. Diversity panels used in genome-wide association studies or multi-parental populations can easily harbour multiple QTL alleles at each locus, more so in the case of polyploids that carry more than two alleles per individual. In such situations a multiallelic model would be closer to reality, allowing for different genetic effects for each potential allele in the population. To obtain such multiallelic markers we propose the usage of haplotypes, concatenations of nearby SNPs. We developed “mpQTL” an R package that can perform a QTL analysis at any ploidy level under biallelic and multiallelic models, depending on the marker type given. We tested the effect of genetic diversity on the power and accuracy difference between bi-allelic and multiallelic models using a set of simulated multiparental autotetraploid, outbreeding populations. Multiallelic models had higher detection power and were more precise than biallelic, SNP-based models, particularly when genetic diversity was higher. This confirms that moving to multi-allelic QTL models can lead to improved detection and characterization of QTLs.

Key message

QTL detection in populations with more than two functional QTL alleles (which is likely in multiparental and/or polyploid populations) is more powerful when using multiallelic models, rather than biallelic models.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-022-04607-z.

The influence of QTL allelic diversity on QTL detection in multi-parent populations: a simulation study in sugar beet

BACKGROUND

Multi-parent populations (MPPs) are important resources for studying plant genetic architecture and detecting quantitative trait loci (QTLs). In MPPs, the QTL effects can show various levels of allelic diversity, which can be an important factor influencing the detection of QTLs. In MPPs, the allelic effects can be more or less specific. They can depend on an ancestor, a parent or the combination of parents in a cross. In this paper, we evaluated the effect of QTL allelic diversity on the QTL detection power in MPPs.

RESULTS

We simulated: a) cross-specific QTLs; b) parental and ancestral QTLs; and c) bi-allelic QTLs. Inspired by a real application in sugar beet, we tested different MPP designs (diallel, chessboard, factorial, and NAM) derived from five or nine parents to explore the ability to sample genetic diversity and detect QTLs. Using a fixed total population size, the QTL detection power was larger in MPPs with fewer but larger crosses derived from a reduced number of parents. The use of a larger set of parents was useful to detect rare alleles with a large phenotypic effect. The benefit of using a larger set of parents was however conditioned on an increase of the total population size. We also determined empirical confidence intervals for QTL location to compare the resolution of different designs. For QTLs representing 6% of the phenotypic variation, using 1600 F₂ offspring individuals, we found average 95% confidence intervals over different designs of 49 and 25 cM for cross-specific and bi-allelic QTLs, respectively.

CONCLUSIONS

MPPs derived from less parents with few but large crosses generally increased the QTL detection power. Using a larger set of parents to cover a wider genetic diversity can be useful to detect QTLs with a reduced minor allele frequency when the QTL effect is large and when the total population size is increased.

Multi-parent multi-environment QTL analysis: an illustration with the EU-NAM Flint population

Multi-parent populations multi-environment QTL experiments data should be analysed jointly to estimate the QTL effect variation within the population and between environments. Commonly, QTL detection in multi-parent populations (MPPs) data measured in multiple environments (ME) is done by analyzing genotypic values 'averaged' across environments. This method ignores the environment-specific QTL (QTLxE) effects. Running separate single environment analyses is a possibility to measure QTLxE effects, but those analyses do not model the genetic covariance due to the use of the same genotype in different environments. In this paper, we propose methods to analyse MPP-ME QTL experiments using simultaneously the data from several environments and modelling the genotypic covariance. Using data from the EU-NAM Flint population, we show that these methods estimate the QTLxE effects and that they can improve the quality of the QTL detection. Those methods also have a larger inference power. For example, they can be extended to integrate environmental indices like temperature or precipitation to better understand the mechanisms behind the QTLxE effects. Therefore, our methodology allows the exploitation of the full MPP-ME data potential: to estimate QTL effect variation (a) within the MPP between sub-populations due to different genetic backgrounds and (b) between environments.

Linkage Analysis and Association Mapping QTL Detection Models for Hybrids Between Multiparental Populations from Two Heterotic Groups: Application to Biomass Production in Maize (Zea mays L.)

Identification of quantitative trait loci (QTL) involved in the variation of hybrid value is of key importance for cross-pollinated species such as maize (Zea mays L.). In a companion paper, we illustrated a new QTL mapping population design involving a factorial mating between two multiparental segregating populations. Six biparental line populations were developed from four founder lines in the Dent and Flint heterotic groups. They were crossed to produce 951 hybrids and evaluated for silage performances. Previously, a linkage analysis (LA) model that assumes each founder line carries a different allele was used to detect QTL involved in General and Specific Combining Abilities (GCA and SCA, respectively) of hybrid value. This previously introduced model requires the estimation of numerous effects per locus, potentially affecting QTL detection power. Using the same design, we compared this “Founder alleles” model to two more parsimonious models, which assume that (i) identity in state at SNP alleles from the same heterotic group implies identity by descent (IBD) at linked QTL (“SNP within-group” model) or (ii) identity in state implies IBD, regardless of population origin of the alleles (“Hybrid genotype” model). This last model assumes biallelic QTL with equal effects in each group. It detected more QTL on average than the two other models but explained lower percentages of variance. The “SNP within-group” model appeared to be a good compromise between the two other models. These results confirm the divergence between the Dent and Flint groups. They also illustrate the need to adapt the QTL detection model to the complexity of the allelic variation, which depends on the trait, the QTL, and the divergence between the heterotic groups.

Reciprocal Genetics: Identifying QTL for General and Specific Combining Abilities in Hybrids Between Multiparental Populations from Two Maize (Zea mays L.) Heterotic Groups

Several plant and animal species of agricultural importance are commercialized as hybrids to take advantage of the heterosis phenomenon. Understanding the genetic architecture of hybrid performances is therefore of key importance. We developed two multiparental maize (Zea mays L.) populations, each corresponding to an important heterotic group (dent or flint) and comprised of six connected biparental segregating populations of inbred lines (802 and 822 lines for each group, respectively) issued from four founder lines. Instead of using "testers" to evaluate their hybrid values, segregating lines were crossed according to an incomplete factorial design to produce 951 dent-flint hybrids, evaluated for four biomass production traits in eight environments. QTL detection was carried out for the general-combining-ability (GCA) and specific-combining-ability (SCA) components of hybrid value, considering allelic effects transmitted from each founder line. In total, 42 QTL were detected across traits. We detected mostly QTL affecting GCA, 31% (41% for dry matter yield) of which also had mild effects on SCA. The small impact of dominant effects is consistent with the known differentiation between the dent and flint heterotic groups and the small percentage of hybrid variance due to SCA observed in our design (∼20% for the different traits). Furthermore, most (80%) of GCA QTL were segregating in only one of the two heterotic groups. Relative to tester-based designs, use of hybrids between two multiparental populations appears highly cost efficient to detect QTL in two heterotic groups simultaneously. This presents new prospects for selecting superior hybrid combinations with markers.

How do the type of QTL effect and the form of the residual term influence QTL detection in multi-parent populations? A case study in the maize EU-NAM population

In the QTL analysis of multi-parent populations, the inclusion of QTLs with various types of effects can lead to a better description of the phenotypic variation and increased power. For the type of QTL effect in QTL models for multi-parent populations (MPPs), various options exist to define them with respect to their origin. They can be modelled as referring to close parental lines or to further away ancestral founder lines. QTL models for MPPs can also be characterized by the homo- or heterogeneity of variance for polygenic effects. The most suitable model for the origin of the QTL effect and the homo- or heterogeneity of polygenic effects may be a function of the genetic distance distribution between the parents of MPPs. We investigated the statistical properties of various QTL detection models for MPPs taking into account the genetic distances between the parents of the MPP. We evaluated models with different assumptions about the QTL effect and the form of the residual term using cross validation. For the EU-NAM data, we showed that it can be useful to mix in the same model QTLs with different types of effects (parental, ancestral, or bi-allelic). The benefit of using cross-specific residual terms to handle the heterogeneity of variance was less obvious for this particular data set.

Genome-wide Association Studies in Maize: Praise and Stargaze

Genome-wide association study (GWAS) has become a widely accepted strategy for decoding genotype-phenotype associations in many species thanks to advances in next-generation sequencing (NGS) technologies. Maize is an ideal crop for GWAS and significant progress has been made in the last decade. This review summarizes current GWAS efforts in maize functional genomics research and discusses future prospects in the omics era. The general goal of GWAS is to link genotypic variations to corresponding differences in phenotype using the most appropriate statistical model in a given population. The current review also presents perspectives for optimizing GWAS design and analysis. GWAS analysis of data from RNA, protein, and metabolite-based omics studies is discussed, along with new models and new population designs that will identify causes of phenotypic variation that have been hidden to date. The joint and continuous efforts of the whole community will enhance our understanding of maize quantitative traits and boost crop molecular breeding designs.

Estimating parent-specific QTL effects through cumulating linked identity-by-state SNP effects in multiparental populations

The emergence of multiparental mapping populations enabled plant geneticists to gain deeper insights into the genetic architecture of major agronomic traits and to map quantitative trait loci (QTLs) controlling the expression of these traits. Although the investigated mapping populations are similar, one open question is whether genotype data should be modelled as identical by state (IBS) or identical by descent (IBD). Whereas IBS simply makes use of raw genotype scores to distinguish alleles, IBD data are derived from parental offspring information. We report on comparing IBS and IBD by applying two multiple regression models on four traits studied in the barley nested association mapping (NAM) population HEB-25. We observed that modelling parent-specific IBD genotypes produced a lower number of significant QTLs with increased prediction abilities compared with modelling IBS genotypes. However, at lower trait heritabilities the IBS model produced higher prediction abilities. We developed a method to estimate multiallelic QTL effects in multiparental populations from simple biallelic IBS data. This method is based on cumulating IBS-derived single-nucleotide polymorphism (SNP) effect estimates in a defined genetic region surrounding a QTL. Comparing the resulting parent-specific QTL effects with those obtained from IBD approaches revealed high accordance that could be confirmed through simulations. The method turned out to be also applicable to a barley multiparent advanced generation inter-cross (MAGIC) population. The 'cumulation method' represents a universal approach to differentiate parent-specific QTL effects in multiparental populations, even if no IBD information is available. In future, the method could further benefit from the availability of much denser SNP maps.

Choice of models for QTL mapping with multiple families and design of the training set for prediction of Fusarium resistance traits in maize

KEY MESSAGE

QTL analysis for Fusarium resistance traits with multiple connected families detected more QTL than single-family analysis. Prediction accuracy was tightly associated with the kinship of the validation and training set.

ABSTRACT

QTL mapping has recently shifted from analysis of single families to multiple, connected families and several biometric models have been suggested. Using a high-density consensus map with 2472 marker loci, we performed QTL mapping with five connected bi-parental families with 639 doubled-haploid (DH) lines in maize for ear rot resistance and analyzed traits DON, Gibberella ear rot severity (GER), and days to silking (DS). Five biometric models differing in the assumption about the number and effects of alleles at QTL were compared. Model 2 to 5 performing joint analyses across all families and using linkage and/or linkage disequilibrium (LD) information identified all and even further QTL than Model 1 (single-family analyses) and generally explained a higher proportion pG of the genotypic variance for all three traits. QTL for DON and GER were mostly family specific, but several QTL for DS occurred in multiple families. Many QTL displayed large additive effects and most alleles increasing resistance originated from a resistant parent. Interactions between detected QTL and genetic background (family) occurred rarely and were comparatively small. Detailed analysis of three fully connected families yielded higher pG values for Model 3 or 4 than for Model 2 and 5, irrespective of the size NTS of the training set (TS). In conclusion, Model 3 and 4 can be recommended for QTL-based prediction with larger families. Including a sufficiently large number of full sibs in the TS helped to increase QTL-based prediction accuracy (rVS) for various scenarios differing in the composition of the TS.

MAGIC populations in crops: current status and future prospects

MAGIC populations present novel challenges and opportunities in crops due to their complex pedigree structure. They offer great potential both for dissecting genomic structure and for improving breeding populations. The past decade has seen the rise of multiparental populations as a study design offering great advantages for genetic studies in plants. The genetic diversity of multiple parents, recombined over several generations, generates a genetic resource population with large phenotypic diversity suitable for high-resolution trait mapping. While there are many variations on the general design, this review focuses on populations where the parents have all been inter-mated, typically termed Multi-parent Advanced Generation Intercrosses (MAGIC). Such populations have already been created in model animals and plants, and are emerging in many crop species. However, there has been little consideration of the full range of factors which create novel challenges for design and analysis in these populations. We will present brief descriptions of large MAGIC crop studies currently in progress to motivate discussion of population construction, efficient experimental design, and genetic analysis in these populations. In addition, we will highlight some recent achievements and discuss the opportunities and advantages to exploit the unique structure of these resources post-QTL analysis for gene discovery.

The crop QTLome comes of age

Recent progress in genomics and phenomics allows for a more accurate and comprehensive characterization of the Quantitative Trait Loci (QTLs)—hereafter defined 'QTLome' as a whole—that govern the variation targeted in breeding programs. High-density genotyping now provides unambiguous identification of QTL alleles, and for several traits beneficial alleles at major QTLs have already been deployed in marker-assisted breeding. However, the amount of QTLome information is enormous and approaches to distill and translate this information to breeders remain to be refined. Improved QTL meta-analyses, better estimation of QTL effects, improved crop modelling and full sharing of raw QTL data will enable a more effective exploitation of the QTLome.

Linkage disequilibrium with linkage analysis of multiline crosses reveals different multiallelic QTL for hybrid performance in the flint and dent heterotic groups of maize

Multiparental designs combined with dense genotyping of parents have been proposed as a way to increase the diversity and resolution of quantitative trait loci (QTL) mapping studies, using methods combining linkage disequilibrium information with linkage analysis (LDLA). Two new nested association mapping designs adapted to European conditions were derived from the complementary dent and flint heterotic groups of maize (Zea mays L.). Ten biparental dent families (N = 841) and 11 biparental flint families (N = 811) were genotyped with 56,110 single nucleotide polymorphism markers and evaluated as test crosses with the central line of the reciprocal design for biomass yield, plant height, and precocity. Alleles at candidate QTL were defined as (i) parental alleles, (ii) haplotypic identity by descent, and (iii) single-marker groupings. Between five and 16 QTL were detected depending on the model, trait, and genetic group considered. In the flint design, a major QTL (R(2) = 27%) with pleiotropic effects was detected on chromosome 10, whereas other QTL displayed milder effects (R(2) < 10%). On average, the LDLA models detected more QTL but generally explained lower percentages of variance, consistent with the fact that most QTL display complex allelic series. Only 15% of the QTL were common to the two designs. A joint analysis of the two designs detected between 15 and 21 QTL for the five traits. Of these, between 27 for silking date and 41% for tasseling date were significant in both groups. Favorable allelic effects detected in both groups open perspectives for improving biomass production.

Combined linkage and linkage disequilibrium QTL mapping in multiple families of maize (Zea mays L.) line crosses highlights complementarities between models based on parental haplotype and single locus polymorphism

Advancements in genotyping are rapidly decreasing marker costs and increasing marker density. This opens new possibilities for mapping quantitative trait loci (QTL), in particular by combining linkage disequilibrium information and linkage analysis (LDLA). In this study, we compared different approaches to detect QTL for four traits of agronomical importance in two large multi-parental datasets of maize (Zea mays L.) of 895 and 928 testcross progenies composed of 7 and 21 biparental families, respectively, and genotyped with 491 markers. We compared to traditional linkage-based methods two LDLA models relying on the dense genotyping of parental lines with 17,728 SNP: one based on a clustering approach of parental line segments into ancestral alleles and one based on single marker information. The two LDLA models generally identified more QTL (60 and 52 QTL in total) than classical linkage models (49 and 44 QTL in total). However, they performed inconsistently over datasets and traits suggesting that a compromise must be found between the reduction of allele number for increasing statistical power and the adequacy of the model to potentially complex allelic variation. For some QTL, the model exclusively based on linkage analysis, which assumed that each parental line carried a different QTL allele, was able to capture remaining variation not explained by LDLA models. These complementarities between models clearly suggest that the different QTL mapping approaches must be considered to capture the different levels of allelic variation at QTL involved in complex traits.