Genomic dissection of variation in clutch size and egg mass in a wild great tit (Parus major) population

Clutch size and egg mass are life history traits that have been extensively studied in wild bird populations, as life history theory predicts a negative trade‐off between them, either at the phenotypic or at the genetic level. Here, we analyse the genomic architecture of these heritable traits in a wild great tit (Parus major) population, using three marker‐based approaches – chromosome partitioning, quantitative trait locus (QTL) mapping and a genome‐wide association study (GWAS). The variance explained by each great tit chromosome scales with predicted chromosome size, no location in the genome contains genome‐wide significant QTL, and no individual SNPs are associated with a large proportion of phenotypic variation, all of which may suggest that variation in both traits is due to many loci of small effect, located across the genome. There is no evidence that any regions of the genome contribute significantly to both traits, which combined with a small, nonsignificant negative genetic covariance between the traits, suggests the absence of genetic constraints on the independent evolution of these traits. Our findings support the hypothesis that variation in life history traits in natural populations is likely to be determined by many loci of small effect spread throughout the genome, which are subject to continued input of variation by mutation and migration, although we cannot exclude the possibility of an additional input of major effect genes influencing either trait.     

Inferences of genetic architecture of bill morphology in house sparrow using a high‐density SNP array point to a polygenic basis

Understanding the genetic architecture of quantitative traits can provide insights into the mechanisms driving phenotypic evolution. Bill morphology is an ecologically important and phenotypically variable trait, which is highly heritable and closely linked to individual fitness. Thus, bill morphology traits are suitable candidates for gene mapping analyses. Previous studies have revealed several genes that may influence bill morphology, but the similarity of gene and allele effects between species and populations is unknown. Here, we develop a custom 200K SNP array and use it to examine the genetic basis of bill morphology in 1857 house sparrow individuals from a large‐scale, island metapopulation off the coast of Northern Norway. We found high genomic heritabilities for bill depth and length, which were comparable with previous pedigree estimates. Candidate gene and genomewide association analyses yielded six significant loci, four of which have previously been associated with craniofacial development. Three of these loci are involved in bone morphogenic protein (BMP) signalling, suggesting a role for BMP genes in regulating bill morphology. However, these loci individually explain a small amount of variance. In combination with results from genome partitioning analyses, this indicates that bill morphology is a polygenic trait. Any studies of eco‐evolutionary processes in bill morphology are therefore dependent on methods that can accommodate polygenic inheritance of the phenotype and molecular‐scale evolution of genetic architecture.     

Genetic architecture of complex traits and accuracy of genomic prediction: coat colour, milk-fat percentage, and type in Holstein cattle as contrasting model traits

Prediction of genetic merit using dense SNP genotypes can be used for estimation of breeding values for selection of livestock, crops, and forage species; for prediction of disease risk; and for forensics. The accuracy of these genomic predictions depends in part on the genetic architecture of the trait, in particular number of loci affecting the trait and distribution of their effects. Here we investigate the difference among three traits in distribution of effects and the consequences for the accuracy of genomic predictions. Proportion of black coat colour in Holstein cattle was used as one model complex trait. Three loci, KIT, MITF, and a locus on chromosome 8, together explain 24% of the variation of proportion of black. However, a surprisingly large number of loci of small effect are necessary to capture the remaining variation. A second trait, fat concentration in milk, had one locus of large effect and a host of loci with very small effects. Both these distributions of effects were in contrast to that for a third trait, an index of scores for a number of aspects of cow confirmation ("overall type"), which had only loci of small effect. The differences in distribution of effects among the three traits were quantified by estimating the distribution of variance explained by chromosome segments containing 50 SNPs. This approach was taken to account for the imperfect linkage disequilibrium between the SNPs and the QTL affecting the traits. We also show that the accuracy of predicting genetic values is higher for traits with a proportion of large effects (proportion black and fat percentage) than for a trait with no loci of large effect (overall type), provided the method of analysis takes advantage of the distribution of loci effects.     

Heritability estimates from genomewide relatedness matrices in wild populations: Application to a passerine,using a small sample size

Genomic developments have empowered the investigation of heritability in wild populations directly from genomewide relatedness matrices (GRM). Such GRM‐based approaches can in particular be used to improve or substitute approaches based on social pedigree (PED‐social). However, measuring heritability from GRM in the wild has not been widely applied yet, especially using small samples and in nonmodel species. Here, we estimated heritability for four quantitative traits (tarsus length, wing length, bill length and body mass), using PED‐social, a pedigree corrected by genetic data (PED‐corrected) and a GRM from a small sample (n = 494) of blue tits from natural populations in Corsica genotyped at nearly 50,000 filtered SNPs derived from RAD‐seq. We also measured genetic correlations among traits, and we performed chromosome partitioning. Heritability estimates were slightly higher when using GRM compared to PED‐social, and PED‐corrected yielded intermediate values, suggesting a minor underestimation of heritability in PED‐social due to incorrect pedigree links, including extra‐pair paternity, and to lower information content than the GRM. Genetic correlations among traits were similar between PED‐social and GRM but credible intervals were very large in both cases, suggesting a lack of power for this small data set. Although a positive linear relationship was found between the number of genes per chromosome and the chromosome heritability for tarsus length, chromosome partitioning similarly showed a lack of power for the three other traits. We discuss the usefulness and limitations of the quantitative genetic inferences based on genomic data in small samples from wild populations.     

Heterogeneity of genetic architecture of body size traits in a free‐living population

Knowledge of the underlying genetic architecture of quantitative traits could aid in understanding how they evolve. In wild populations, it is still largely unknown whether complex traits are polygenic or influenced by few loci with major effect, due to often small sample sizes and low resolution of marker panels. Here, we examine the genetic architecture of five adult body size traits in a free‐living population of Soay sheep on St Kilda using 37 037 polymorphic SNPs. Two traits (jaw and weight) show classical signs of a polygenic trait: the proportion of variance explained by a chromosome was proportional to its length, multiple chromosomes and genomic regions explained significant amounts of phenotypic variance, but no SNPs were associated with trait variance when using GWAS. In comparison, genetic variance for leg length traits (foreleg, hindleg and metacarpal) was disproportionately explained by two SNPs on chromosomes 16 (s23172.1) and 19 (s74894.1), which each explained >10% of the additive genetic variance. After controlling for environmental differences, females heterozygous for s74894.1 produced more lambs and recruits during their lifetime than females homozygous for the common allele conferring long legs. We also demonstrate that alleles conferring shorter legs have likely entered the population through a historic admixture event with the Dunface sheep. In summary, we show that different proxies for body size can have very different genetic architecture and that dense SNP helps in understanding both the mode of selection and the evolutionary history at loci underlying quantitative traits in natural populations.     

Genetic architecture of ovary size and asymmetry in European honeybee workers

The molecular basis of complex traits is increasingly understood but a remaining challenge is to identify their co-regulation and inter-dependence. Pollen hoarding (pln) in honeybees is a complex trait associated with a well-characterized suite of linked behavioral and physiological traits. In European honeybee stocks bidirectionally selected for pln, worker (sterile helper) ovary size is pleiotropically affected by quantitative trait loci that were initially identified for their effect on foraging behavior. To gain a better understanding of the genetic architecture of worker ovary size in this model system, we analyzed a series of crosses between the selected strains. The crossing results were heterogeneous and suggested non-additive effects. Three significant and three suggestive quantitative trait loci of relatively large effect sizes were found in two reciprocal backcrosses. These loci are not located in genome regions of known effects on foraging behavior but contain several interesting candidate genes that may specifically affect worker-ovary size. Thus, the genetic architecture of this life history syndrome may be comprised of pleiotropic, central regulators that influence several linked traits and other genetic factors that may be downstream and trait specific.     

Deciphering the genomic architecture of the stickleback brain with a novel multilocus gene‐mapping approach

Quantitative traits important to organismal function and fitness, such as brain size, are presumably controlled by many small‐effect loci. Deciphering the genetic architecture of such traits with traditional quantitative trait locus (QTL) mapping methods is challenging. Here, we investigated the genetic architecture of brain size (and the size of five different brain parts) in nine‐spined sticklebacks (Pungitius pungitius) with the aid of novel multilocus QTL‐mapping approaches based on a de‐biased LASSO method. Apart from having more statistical power to detect QTL and reduced rate of false positives than conventional QTL‐mapping approaches, the developed methods can handle large marker panels and provide estimates of genomic heritability. Single‐locus analyses of an F₂ interpopulation cross with 239 individuals and 15 198, fully informative single nucleotide polymorphisms (SNPs) uncovered 79 QTL associated with variation in stickleback brain size traits. Many of these loci were in strong linkage disequilibrium (LD) with each other, and consequently, a multilocus mapping of individual SNPs, accounting for LD structure in the data, recovered only four significant QTL. However, a multilocus mapping of SNPs grouped by linkage group (LG) identified 14 LGs (1–6 depending on the trait) that influence variation in brain traits. For instance, 17.6% of the variation in relative brain size was explainable by cumulative effects of SNPs distributed over six LGs, whereas 42% of the variation was accounted for by all 21 LGs. Hence, the results suggest that variation in stickleback brain traits is influenced by many small‐effect loci. Apart from suggesting moderately heritable (h² ≈ 0.15–0.42) multifactorial genetic architecture of brain traits, the results highlight the challenges in identifying the loci contributing to variation in quantitative traits. Nevertheless, the results demonstrate that the novel QTL‐mapping approach developed here has distinctive advantages over the traditional QTL‐mapping methods in analyses of dense marker panels.     

Selection strategies for linkage studies using twins.

Genetic linkage analysis for complex diseases offers a major challenge to geneticists. In these complex diseases multiple genetic loci are responsible for the disease and they may vary in the size of their contribution; the effect of any single one of them is likely to be small. In many situations, like in extensive twin registries, trait values have been recorded for a large number of individuals, and preliminary studies have revealed summary measures for those traits, like mean, variance and components of variance, including heritability. Given the small effect size, a random sample of twins will require a prohibitively large sample size. It is well known that selective sampling is far more efficient in terms of genotyping effort. In this paper we derive easy expressions for the information contributed by sib pairs for the detection of linkage to a quantitative trait locus (QTL). We consider random samples as well as samples of sib pairs selected on the basis of their trait values. These expressions can be rapidly computed and do not involve simulation. We extend our results for quantitative traits to dichotomous traits using the concept of a liability threshold model. We present tables with required sample sizes for height, insulin levels and migraine, three of the traits studied in the GenomEUtwin project.     

Simultaneous Discovery,Estimation and Prediction Analysis of Complex Traits Using a Bayesian Mixture Model

Gene discovery, estimation of heritability captured by SNP arrays, inference on genetic architecture and prediction analyses of complex traits are usually performed using different statistical models and methods, leading to inefficiency and loss of power. Here we use a Bayesian mixture model that simultaneously allows variant discovery, estimation of genetic variance explained by all variants and prediction of unobserved phenotypes in new samples. We apply the method to simulated data of quantitative traits and Welcome Trust Case Control Consortium (WTCCC) data on disease and show that it provides accurate estimates of SNP-based heritability, produces unbiased estimators of risk in new samples, and that it can estimate genetic architecture by partitioning variation across hundreds to thousands of SNPs. We estimated that, depending on the trait, 2,633 to 9,411 SNPs explain all of the SNP-based heritability in the WTCCC diseases. The majority of those SNPs (>96%) had small effects, confirming a substantial polygenic component to common diseases. The proportion of the SNP-based variance explained by large effects (each SNP explaining 1% of the variance) varied markedly between diseases, ranging from almost zero for bipolar disorder to 72% for type 1 diabetes. Prediction analyses demonstrate that for diseases with major loci, such as type 1 diabetes and rheumatoid arthritis, Bayesian methods outperform profile scoring or mixed model approaches.     

Genetic architecture of the dog: sexual size dimorphism and functional morphology

Purebred dogs are a valuable resource for genetic analysis of quantitative traits. Quantitative traits are complex, controlled by many genes that are contained within regions of the genome known as quantitative trait loci (QTL). The genetic architecture of quantitative traits is defined by the characteristics of these genes: their number, the magnitude of their effects, their positions in the genome and their interactions with each other. QTL analysis is a valuable tool for exploring genetic architecture, and highlighting regions of the genome that contribute to the variation of a trait within a population.     

Epistasis affecting litter size in mice

Litter size is an important reproductive trait as it makes a major contribution to fitness. Generally, traits closely related to fitness show low heritability perhaps because of the corrosive effects of directional natural selection on the additive genetic variance. Nonetheless, low heritability does not imply, necessarily, a complete absence of genetic variation because genetic interactions (epistasis and dominance) contribute to variation in traits displaying strong heterosis in crosses, such as litter size. In our study, we investigated the genetic architecture of litter size in 166 females from an F2 intercross of the SM/J and LG/J inbred mouse strains. Litter size had a low heritability (h2 = 12%) and a low repeatability (r = 33%). Using interval-mapping methods, we located two quantitative trait loci (QTL) affecting litter size at locations D7Mit21 + 0 cM and D12Mit6 + 8 cM, on chromosomes 7 and 12 respectively. These QTL accounted for 12.6% of the variance in litter size. In a two-way genome-wide epistasis scan we found eight QTL interacting epistatically involving chromosomes 2, 4, 5, 11, 14, 15 and 18. Taken together, the QTL and their interactions explain nearly 49% (39.5% adjusted multiple r2) of the phenotypic variation for litter size in this cross, an increase of 36% over the direct effects of the QTL. This indicates the importance of epistasis as a component of the genetic architecture of litter size and fitness in our intercross population.     

Presence of multiple independent effects in risk loci of common complex human diseases

Many genetic loci and SNPs associated with many common complex human diseases and traits are now identified. The total genetic variance explained by these loci for a trait or disease, however, has often been very small. Much of the "missing heritability" has been revealed to be hidden in the genome among the large number of variants with small effects. Several recent studies have reported the presence of multiple independent SNPs and genetic heterogeneity in trait-associated loci. It is therefore reasonable to speculate that such a phenomenon could be common among loci known to be associated with a complex trait or disease. For testing this hypothesis, a total of 117 loci known to be associated with rheumatoid arthritis (RA), Crohn disease (CD), type 1 diabetes (T1D), or type 2 diabetes (T2D) were selected. The presence of multiple independent effects was assessed in the case-control samples genotyped by the Wellcome Trust Case Control Consortium study and imputed with SNP genotype information from the HapMap Project and the 1000 Genomes Project. Eleven loci with evidence of multiple independent effects were identified in the study, and the number was expected to increase at larger sample sizes and improved statistical power. The variance explained by the multiple effects in a locus was much higher than the variance explained by the single reported SNP effect. The results thus significantly improve our understanding of the allelic structure of these individual disease-associated loci, as well as our knowledge of the general genetic mechanisms of common complex traits and diseases.     

Comparison of whole-genome prediction models for traits with contrasting genetic architecture in a diversity panel of maize inbred lines

ABSTRACT: BACKGROUND: There is increasing empirical evidence that whole-genome prediction (WGP) is a powerful tool for predicting line and hybrid performance in maize. However, there is a lack of knowledge about the sensitivity of WGP models towards the genetic architecture of the trait. Whereas previous studies exclusively focused on highly polygenic traits, important agronomic traits such as disease resistances, nutrifunctional or climate adaptational traits have a genetic architecture which is either much less complex or unknown. For such cases, information about model robustness and guidelines for model selection are lacking. Here, we compared five WGP models with different assumptions about the distribution of the underlying genetic effects. As contrasting model traits, we chose three highly polygenic agronomic traits and three metabolites each with a major QTL explaining 22 to 30 % of the genetic variance in a panel of 289 diverse maize inbred lines genotyped with 56,110 SNPs. RESULTS: We found the five WGP models to be remarkable robust towards trait architecture with the largest differences in prediction accuracies ranging between 0.05 and 0.14 for the same trait, most likely as the result of the high level of linkage disequilibrium prevailing in elite maize germplasm. Whereas RR-BLUP performed best for the agronomic traits, it was inferior to LASSO or elastic net for the three metabolites. We found the approach of genome partitioning of genetic variance, first applied in human genetics, as useful in guiding the breeder which model to choose, if prior knowledge of the trait architecture is lacking. CONCLUSIONS: Our results suggest that in diverse germplasm of elite maize inbred lines with a high level of LD, WGP models differ only slightly in their accuracies, irrespective of the number and effects of QTL found in previous linkage or association mapping studies. However, small gains in prediction accuracies can be achieved if the WGP model is selected according to the genetic architecture of the trait. If the trait architecture is unknown e.g. for novel traits which only recently received attention in breeding, we suggest to inspect the distribution of the genetic variance explained by each chromosome for guiding model selection in WGP.     

Determination of the genetic architecture of seed size and shape via linkage and association analysis in soybean (Glycine max L. Merr.)

Seed-size traits, which are controlled by multiple genes in soybean, play an important role in determining seed yield, quality and appearance. However, the molecular mechanisms controlling the size of soybean seeds remain unclear, and little research has been done to investigate these mechanisms. In this study, we performed a genetic analysis to determine the genetic architecture of soybean seed size and shape via linkage and association analyses. We used 184 recombinant inbred lines (RILs) and 219 cultivated soybean accessions to evaluate seed length, seed width and seed height as seed-size traits, and their ratios of these values as seed-shape traits. Our results showed that all six traits had high heritability ranging from 92.46 to 98.47 %. Linkage analysis in the RILs identified 12 quantitative traits loci (QTLs), with five of these QTLs being associated with seed size, five with seed shape and two with the two first principal components of our principal component analysis (PCA). Association analysis in the 219 accessions detected 41 single nucleotide polymorphism (SNP)-trait associations, with 20 of these SNPs being associated with seed-size traits, seven with seed-shape traits and 14 with the two first principal components of our PCA. This analysis reveals that seed-size and seed-shape may be controlled by different genetic factors. Our results provide a greater understanding of phenotypic structure and genetic architecture of soybean seed, and the QTLs detected in this study form a basis for future fine mapping, quantitative trait gene cloning and molecular breeding in soybean.     

Genome partitioning of genetic variation for height from 11,214 sibling pairs

Height has been used for more than a century as a model by which to understand quantitative genetic variation in humans. We report that the entire genome appears to contribute to its additive genetic variance. We used genotypes and phenotypes of 11,214 sibling pairs from three countries to partition additive genetic variance across the genome. Using genome scans to estimate the proportion of the genomes of each chromosome from siblings that were identical by descent, we estimated the heritability of height contributed by each of the 22 autosomes and the X chromosome. We show that additive genetic variance is spread across multiple chromosomes and that at least six chromosomes (i.e., 3, 4, 8, 15, 17, and 18) are responsible for the observed variation. Indeed, the data are not inconsistent with a uniform spread of trait loci throughout the genome. Our estimate of the variance explained by a chromosome is correlated with the number of times suggestive or significant linkage with height has been reported for that chromosome. Variance due to dominance was not significant but was difficult to assess because of the high sampling correlation between additive and dominance components. Results were consistent with the absence of any large between-chromosome epistatic effects. Notwithstanding the proposed architecture of complex traits that involves widespread gene-gene and gene-environment interactions, our results suggest that variation in height in humans can be explained by many loci distributed over all autosomes, with an additive mode of gene action.     

Replicated analysis of the genetic architecture of quantitative traits in two wild great tit populations

Currently, there is much debate on the genetic architecture of quantitative traits in wild populations. Is trait variation influenced by many genes of small effect or by a few genes of major effect? Where is additive genetic variation located in the genome? Do the same loci cause similar phenotypic variation in different populations? Great tits (Parus major) have been studied extensively in long‐term studies across Europe and consequently are considered an ecological ‘model organism’. Recently, genomic resources have been developed for the great tit, including a custom SNP chip and genetic linkage map. In this study, we used a suite of approaches to investigate the genetic architecture of eight quantitative traits in two long‐term study populations of great tits—one in the Netherlands and the other in the United Kingdom. Overall, we found little evidence for the presence of genes of large effects in either population. Instead, traits appeared to be influenced by many genes of small effect, with conservative estimates of the number of contributing loci ranging from 31 to 310. Despite concordance between population‐specific heritabilities, we found no evidence for the presence of loci having similar effects in both populations. While population‐specific genetic architectures are possible, an undetected shared architecture cannot be rejected because of limited power to map loci of small and moderate effects. This study is one of few examples of genetic architecture analysis in replicated wild populations and highlights some of the challenges and limitations researchers will face when attempting similar molecular quantitative genetic studies in free‐living populations.     

Are we underestimating the genetic variances of dimorphic traits?

Populations often contain discrete classes or morphs (e.g., sexual dimorphisms, wing dimorphisms, trophic dimorphisms) characterized by distinct patterns of trait expression. In quantitative genetic analyses, the different morphs can be considered as different environments within which traits are expressed. Genetic variances and covariances can then be estimated independently for each morph or in a combined analysis. In the latter case, morphs can be considered as separate environments in a bivariate analysis or entered as fixed effects in a univariate analysis. Although a common approach, we demonstrate that the latter produces downwardly biased estimates of additive genetic variance and heritability unless the quantitative genetic architecture of the traits concerned is perfectly correlated between the morphs. This result is derived for four widely used quantitative genetic variance partitioning methods. Given that theory predicts the evolution of genotype‐by‐environment (morph) interactions as a consequence of selection favoring different trait combinations in each morph, we argue that perfect correlations between the genetic architectures of the different morphs are unlikely. A sampling of the recent literature indicates that the majority of researchers studying traits expressed in different morphs recognize this and do estimate morph‐specific quantitative genetic architecture. However, ca. 16% of the studies in our sample utilized only univariate, fixed‐effects models. We caution against this approach and recommend that it be used only if supported by evidence that the genetic architectures of the different morphs do not differ.     

Sex-specific genetic architecture of human fatness in Chinese: the SAPPHIRe Study

To dissect the genetic architecture of sexual dimorphism in obesity-related traits, we evaluated the sex–genotype interaction, sex-specific heritability and genome-wide linkages for seven measurements related to obesity. A total of 1,365 non-diabetic Chinese subjects from the family study of the Stanford Asia–Pacific Program of Hypertension and Insulin Resistance were used to search for quantitative trait loci (QTLs) responsible for the obesity-related traits. Pleiotropy and co-incidence effects from the QTLs were also examined using the bivariate linkage approach. We found that sex-specific differences in heritability and the genotype–sex interaction effects were substantially significant for most of these traits. Several QTLs with strong linkage evidence were identified after incorporating genotype by sex (G × S) interactions into the linkage mapping, including one QTL for hip circumference [maximum LOD score (MLS) = 4.22, empirical p = 0.000033] and two QTLs: for BMI on chromosome 12q with MLS 3.37 (empirical p = 0.0043) and 3.10 (empirical p = 0.0054). Sex-specific analyses demonstrated that these linkage signals all resulted from females rather than males. Most of these QTLs for obesity-related traits replicated the findings in other ethnic groups. Bivariate linkage analyses showed several obesity traits were influenced by a common set of QTLs. All regions with linkage signals were observed in one gender, but not in the whole sample, suggesting the genetic architecture of obesity-related traits does differ by gender. These findings are useful for further identification of the liability genes for these phenotypes through candidate genes or genome-wide association analysis.     

Genome-Wide Prediction of Traits with Different Genetic Architecture Through Efficient Variable Selection

In genome-based prediction there is considerable uncertainty about the statistical model and method required to maximize prediction accuracy. For traits influenced by a small number of quantitative trait loci (QTL), predictions are expected to benefit from methods performing variable selection [e.g., BayesB or the least absolute shrinkage and selection operator (LASSO)] compared to methods distributing effects across the genome [ridge regression best linear unbiased prediction (RR-BLUP)]. We investigate the assumptions underlying successful variable selection by combining computer simulations with large-scale experimental data sets from rice (Oryza sativa L.), wheat (Triticum aestivum L.), and Arabidopsis thaliana (L.). We demonstrate that variable selection can be successful when the number of phenotyped individuals is much larger than the number of causal mutations contributing to the trait. We show that the sample size required for efficient variable selection increases dramatically with decreasing trait heritabilities and increasing extent of linkage disequilibrium (LD). We contrast and discuss contradictory results from simulation and experimental studies with respect to superiority of variable selection methods over RR-BLUP. Our results demonstrate that due to long-range LD, medium heritabilities, and small sample sizes, superiority of variable selection methods cannot be expected in plant breeding populations even for traits like FRIGIDA gene expression in Arabidopsis and flowering time in rice, assumed to be influenced by a few major QTL. We extend our conclusions to the analysis of whole-genome sequence data and infer upper bounds for the number of causal mutations which can be identified by LASSO. Our results have major impact on the choice of statistical method needed to make credible inferences about genetic architecture and prediction accuracy of complex traits.