Beyond missing heritability: prediction of complex traits

Despite rapid advances in genomic technology, our ability to account for phenotypic variation using genetic information remains limited for many traits. This has unfortunately resulted in limited application of genetic data towards preventive and personalized medicine, one of the primary impetuses of genome-wide association studies. Recently, a large proportion of the "missing heritability" for human height was statistically explained by modeling thousands of single nucleotide polymorphisms concurrently. However, it is currently unclear how gains in explained genetic variance will translate to the prediction of yet-to-be observed phenotypes. Using data from the Framingham Heart Study, we explore the genomic prediction of human height in training and validation samples while varying the statistical approach used, the number of SNPs included in the model, the validation scheme, and the number of subjects used to train the model. In our training datasets, we are able to explain a large proportion of the variation in height (h(2) up to 0.83, R(2) up to 0.96). However, the proportion of variance accounted for in validation samples is much smaller (ranging from 0.15 to 0.36 depending on the degree of familial information used in the training dataset). While such R(2) values vastly exceed what has been previously reported using a reduced number of pre-selected markers (<0.10), given the heritability of the trait (～ 0.80), substantial room for improvement remains.     

Estimates of missing heritability for complex traits in Brown Swiss cattle

Background

Genomic selection estimates genetic merit based on dense SNP (single nucleotide polymorphism) genotypes and phenotypes. This requires that SNPs explain a large fraction of the genetic variance. The objectives of this work were: (1) to estimate the fraction of genetic variance explained by dense genome-wide markers using 54 K SNP chip genotyping, and (2) to evaluate the effect of alternative marker-based relationship matrices and corrections for the base population on the fraction of the genetic variance explained by markers.

Methods

Two alternative marker-based relationship matrices were estimated using 35 706 SNPs on 1086 dairy bulls. Both pedigree- and marker-based relationship matrices were fitted simultaneously or separately in an animal model to estimate the fraction of variance not explained by the markers, i.e. the fraction explained by the pedigree. The phenotypes considered in the analysis were the deregressed estimated breeding values (dEBV) for milk, fat and protein yield and for somatic cell score (SCS).

Results

When dEBV were not sufficiently accurate (50 or 70%), the estimated fraction of the genetic variance explained by the markers was around 65% for yield traits and 45% for SCS. Scaling marker genotypes with locus-specific frequencies of heterozygotes slightly increased the variance explained by markers, compared with scaling with the average frequency of heterozygotes across loci. The estimated fraction of the genetic variance explained by the markers using separately both relationships matrices followed the same trends but the results were underestimated. With less accurate dEBV estimates, the fraction of the genetic variance explained by markers was underestimated, which is probably an artifact due to the dEBV being estimated by a pedigree-based animal model.

Conclusions

When using only highly accurate dEBV, the proportion of the genetic variance explained by the Illumina 54 K SNP chip was approximately 80% for Brown Swiss cattle. These results depend on the SNP chip used and the family structure of the population, i.e. more dense SNPs and closer family relationships are expected to result in a higher fraction of the variance explained by the SNPs.     

Parameters in dynamic models of complex traits are containers of missing heritability

Polymorphisms identified in genome-wide association studies of human traits rarely explain more than a small proportion of the heritable variation, and improving this situation within the current paradigm appears daunting. Given a well-validated dynamic model of a complex physiological trait, a substantial part of the underlying genetic variation must manifest as variation in model parameters. These parameters are themselves phenotypic traits. By linking whole-cell phenotypic variation to genetic variation in a computational model of a single heart cell, incorporating genotype-to-parameter maps, we show that genome-wide association studies on parameters reveal much more genetic variation than when using higher-level cellular phenotypes. The results suggest that letting such studies be guided by computational physiology may facilitate a causal understanding of the genotype-to-phenotype map of complex traits, with strong implications for the development of phenomics technology.     

Genome-wide association mapping of blood cell traits in mice

Genetic variations in blood cell parameters can impact clinical traits. We report here the mapping of blood cell traits in a panel of 100 inbred strains of mice of the Hybrid Mouse Diversity Panel (HMDP) using genome-wide association (GWA). We replicated a locus previously identified in using linkage analysis in several genetic crosses for mean corpuscular volume (MCV) and a number of other red blood cell traits on distal chromosome 7. Our peak for SNP association to MCV occurred in a linkage disequilibrium (LD) block spanning from 109.38 to 111.75 Mb that includes Hbb-b1, the likely causal gene. Altogether, we identified five loci controlling red blood cell traits (on chromosomes 1, 7, 11, 12, and 16), and four of these correspond to loci for red blood cell traits reported in a recent human GWA study. For white blood cells, including granulocytes, monocytes, and lymphocytes, a total of six significant loci were identified on chromosomes 1, 6, 8, 11, 12, and 15. An average of ten candidate genes were found at each locus and those were prioritized by examining functional variants in the HMDP such as missense and expression variants. These results provide intermediate phenotypes and candidate loci for genetic studies of atherosclerosis and cancer as well as inflammatory and immune disorders in mice.     

QTL mapping for yield and yield contributing traits in two mapping populations of bread wheat

In bread wheat, single-locus and two-locus QTL analyses were conducted for seven yield and yield contributing traits using two different mapping populations (P I and P II). Single-locus QTL analyses involved composite interval mapping (CIM) for individual traits and multiple-trait composite interval mapping (MCIM) for correlated yield traits to detect the pleiotropic QTLs. Two-locus analyses were conducted to detect main effect QTLs (M-QTLs), epistatic QTLs (E-QTLs) and QTL × environment interactions (QE and QQE). Only a solitary QTL for spikelets per spike was common between the above two populations. HomoeoQTLs were also detected, suggesting the presence of triplicate QTLs in bread wheat. Relatively fewer QTLs were detected in P I than in P II. This may be partly due to low density of marker loci on P I framework map (173) than in P II (521) and partly due to more divergent parents used for developing P II. Six QTLs were important which were pleiotropic/coincident involving more than one trait and were also consistent over environments. These QTLs could be utilized efficiently for marker assisted selection (MAS).     

Effects of missing marker and segregation distortion on QTL mapping in F2 populations

Missing marker and segregation distortion are commonly encountered in actual quantitative trait locus (QTL) mapping populations. Our objective in this study was to investigate the impact of the two factors on QTL mapping through computer simulations. Results indicate that detection power decreases with increasing levels of missing markers, and the false discovery rate increases. Missing markers have greater effects on smaller effect QTL and smaller size populations. The effect of missing markers can be quantified by a population with a reduced size similar to the marker missing rate. As for segregation distortion, if the distorted marker is not closely linked with any QTL, it will not have significant impact on QTL mapping; otherwise, the impact of the distortion will depend on the degree of dominance of QTL, frequencies of the three marker types, the linkage distance between the distorted marker and QTL, and the mapping population size. Sometimes, the distortion can result in a higher genetic variance than that of non-distortion, and therefore benefits the detection of linked QTL. A formula of the ratio of genetic variance explained by QTL under distortion and non-distortion was given in this study, so as to easily determine whether the segregation distortion marker (SDM) increases or decreases the QTL detection power. The effect of SDM decreases rapidly as its linkage relationship with QTL becomes looser. In general, distorted markers will not have a great effect on the position and effect estimations of QTL, and their effects can be ignored in large-size mapping populations.     

Linkage mapping of CVD risk traits in the isolated Norfolk Island population

To understand the underlying genetic architecture of cardiovascular disease (CVD) risk traits, we undertook a genome-wide linkage scan to identify CVD quantitative trait loci (QTLs) in 377 individuals from the Norfolk Island population. The central aim of this research focused on the utilization of a genetically and geographically isolated population of individuals from Norfolk Island for the purposes of variance component linkage analysis to identify QTLs involved in CVD risk traits. Substantial evidence supports the involvement of traits such as systolic and diastolic blood pressures, high-density lipoprotein-cholesterol, low-density lipoprotein-cholesterol, body mass index and triglycerides as important risk factors for CVD pathogenesis. In addition to the environmental influences of poor diet, reduced physical activity, increasing age, cigarette smoking and alcohol consumption, many studies have illustrated a strong involvement of genetic components in the CVD phenotype through family and twin studies. We undertook a genome scan using 400 markers spaced approximately 10 cM in 600 individuals from Norfolk Island. Genotype data was analyzed using the variance components methods of SOLAR. Our results gave a peak LOD score of 2.01 localizing to chromosome 1p36 for systolic blood pressure and replicated previously implicated loci for other CVD relevant QTLs.     

QTLNetwork: mapping and visualizing genetic architecture of complex traits in experimental populations

SUMMARY: QTLNetwork is a software package for mapping and visualizing the genetic architecture underlying complex traits for experimental populations derived from a cross between two inbred lines. It can simultaneously map quantitative trait loci (QTL) with individual effects, epistasis and QTL-environment interaction. Currently, it is able to handle data from F(2), backcross, recombinant inbred lines and double-haploid populations, as well as populations from specific mating designs (immortalized F(2) and BC(n)F(n) populations). The Windows version of QTLNetwork was developed with a graphical user interface. Alternatively, the command-line versions have the facility to be run in other prevalent operating systems, such as Linux, Unix and MacOS. AVAILABILITY: http://ibi.zju.edu.cn/software/qtlnetwork.     

Ocular refraction: heritability and genome-wide search for eye morphometry traits in an isolated Sardinian population

No genes influencing oculometric phenotypes have yet been identified, despite it being well known that eye morphometry is involved in refraction and that genetics may play an important role. We have therefore performed a heritability analysis and genome-wide search (GWS) of biometric ocular traits in an isolated Sardinian population, assessing the genetic contribution and identifying the associated genetic loci. A complete eye examination including refraction and ocular biometry measurements such as axial length (AL), anterior chamber depth (ACD) and corneal curvature (CC), was performed on 789 subjects. Heritability analysis was carried out by means of parent–offspring regression and variance component models. Univariate and bivariate linkage analysis was performed by using 654 microsatellite markers spanning the genome. CC showed a mean heritability of 57%. AL and ACD were found to have significantly different variances (P<0.01) in males and females, so that heritability was calculated separately for each sex. AL had an estimated heritability in females of 31% and in males of 60%, whereas ACD had an estimated heritability of 47% in females and of 44% in males. In the GWS, the most suggestive evidence of linkage was identified on chromosome 2 for AL (LOD 2.64), on chromosome 1 for ACD (LOD 2.32) and on chromosomes 7, 2 and 3 for CC (LOD 2.50, 2.44 and 2.34, respectively). High heritability of eye morphometry traits was thus revealed. The identified loci are the first linkage signals available in ocular biometry. Notably, the observed significant differences in parental transmission deserve further study.The authors Ginevra Biino and Maria Antonietta Palmas contributed equally to this work     

Repeatability and heritability of reproductive traits in free-ranging snakes

The underlying genetic basis of life-history traits in free-ranging animals is critical to the effects of selection on such traits, but logistical constraints mean that such data are rarely available. Our long-term ecological studies on free-ranging oviparous snakes (keelbacks, Tropidonophis mairii (Gray, 1841), Colubridae) on an Australian floodplain provide the first such data for any tropical reptile. All size-corrected reproductive traits (egg mass, clutch size, clutch mass and post-partum maternal mass) were moderately repeatable between pairs of clutches produced by 69 female snakes after intervals of 49-1152 days, perhaps because maternal body condition was similar between clutches. Parent-offspring regression of reproductive traits of 59 pairs of mothers and daughters revealed high heritability for egg mass (h2= 0.73, SE=0.24), whereas heritability for the other three traits was low (< 0.37). The estimated heritability of egg mass may be inflated by maternal effects such as differential allocation of yolk steroids to different-sized eggs. High heritability of egg size may be maintained (rather than eroded by stabilizing selection) because selection acts on a trait (hatchling size) that is determined by the interaction between egg size and incubation substrate rather than by egg size alone. Variation in clutch size was mainly because of environmental factors (h2=0.04), indicating that one component of the trade-off between egg size and clutch size is under much tighter genetic control than the other. Thus, the phenotypic trade-off between egg size and egg number in keelback snakes occurs because each female snake must allocate a finite amount of energy into eggs of a genetically determined size.     

Genome-wide association mapping of agronomic and morphologic traits in highly structured populations of barley cultivars

Genome-wide association study (GWAS) has become an obvious general approach for studying traits of agricultural importance in higher plants, especially crops. Here, we present a GWAS of 32 morphologic and 10 agronomic traits in a collection of 615 barley cultivars genotyped by genome-wide polymorphisms from a recently developed barley oligonucleotide pool assay. Strong population structure effect related to mixed sampling based on seasonal growth habit and ear row number is present in this barley collection. Comparison of seven statistical approaches in a genome-wide scan for significant associations with or without correction for confounding by population structure, revealed that in reducing false positive rates while maintaining statistical power, a mixed linear model solution outperforms genomic control, structured association, stepwise regression control and principal components adjustment. The present study reports significant associations for sixteen morphologic and nine agronomic traits and demonstrates the power and feasibility of applying GWAS to explore complex traits in highly structured plant samples.     

Genomic selection for growth and wood quality in Eucalyptus: capturing the missing heritability and accelerating breeding for complex traits in forest trees

? Genomic selection (GS) is expected to cause a paradigm shift in tree breeding by improving its speed and efficiency. By fitting all the genome-wide markers concurrently, GS can capture most of the 'missing heritability' of complex traits that quantitative trait locus (QTL) and association mapping classically fail to explain. Experimental support of GS is now required. ? The effectiveness of GS was assessed in two unrelated Eucalyptus breeding populations with contrasting effective population sizes (N(e) = 11 and 51) genotyped with > 3000 DArT markers. Prediction models were developed for tree circumference and height growth, wood specific gravity and pulp yield using random regression best linear unbiased predictor (BLUP). ? Accuracies of GS varied between 0.55 and 0.88, matching the accuracies achieved by conventional phenotypic selection. Substantial proportions (74-97%) of trait heritability were captured by fitting all genome-wide markers simultaneously. Genomic regions explaining trait variation largely coincided between populations, although GS models predicted poorly across populations, likely as a result of variable patterns of linkage disequilibrium, inconsistent allelic effects and genotype × environment interaction. ? GS brings a new perspective to the understanding of quantitative trait variation in forest trees and provides a revolutionary tool for applied tree improvement. Nevertheless population-specific predictive models will likely drive the initial applications of GS in forest tree breeding.     

Genome-wide linkage mapping of yield-related traits in three Chinese bread wheat populations using high-density SNP markers

Key message

We identified 21 new and stable QTL, and 11 QTL clusters for yield-related traits in three bread wheat populations using the wheat 90 K SNP assay.

Abstract

Identification of quantitative trait loci (QTL) for yield-related traits and closely linked molecular markers is important in order to identify gene/QTL for marker-assisted selection (MAS) in wheat breeding. The objectives of the present study were to identify QTL for yield-related traits and dissect the relationships among different traits in three wheat recombinant inbred line (RIL) populations derived from crosses Doumai?×?Shi 4185 (D?×?S), Gaocheng 8901?×?Zhoumai 16 (G?×?Z) and Linmai 2?×?Zhong 892 (L?×?Z). Using the available high-density linkage maps previously constructed with the wheat 90 K iSelect single nucleotide polymorphism (SNP) array, 65, 46 and 53 QTL for 12 traits were identified in the three RIL populations, respectively. Among them, 34, 23 and 27 were likely to be new QTL. Eighteen common QTL were detected across two or three populations. Eleven QTL clusters harboring multiple QTL were detected in different populations, and the interval 15.5–32.3 cM around the Rht-B1 locus on chromosome 4BS harboring 20 QTL is an important region determining grain yield (GY). Thousand-kernel weight (TKW) is significantly affected by kernel width and plant height (PH), whereas flag leaf width can be used to select lines with large kernel number per spike. Eleven candidate genes were identified, including eight cloned genes for kernel, heading date (HD) and PH-related traits as well as predicted genes for TKW, spike length and HD. The closest SNP markers of stable QTL or QTL clusters can be used for MAS in wheat breeding using kompetitive allele-specific PCR or semi-thermal asymmetric reverse PCR assays for improvement of GY.

     

Inclusive composite interval mapping (ICIM) for digenic epistasis of quantitative traits in biparental populations

It has long been recognized that epistasis or interactions between non-allelic genes plays an important role in the genetic control and evolution of quantitative traits. However, the detection of epistasis and estimation of epistatic effects are difficult due to the complexity of epistatic patterns, insufficient sample size of mapping populations and lack of efficient statistical methods. Under the assumption of additivity of QTL effects on the phenotype of a trait in interest, the additive effect of a QTL can be completely absorbed by the flanking marker variables, and the epistatic effect between two QTL can be completely absorbed by the four marker-pair multiplication variables between the two pairs of flanking markers. Based on this property, we proposed an inclusive composite interval mapping (ICIM) by simultaneously considering marker variables and marker-pair multiplications in a linear model. Stepwise regression was applied to identify the most significant markers and marker-pair multiplications. Then a two-dimensional scanning (or interval mapping) was conducted to identify QTL with significant digenic epistasis using adjusted phenotypic values based on the best multiple regression model. The adjusted values retain the information of QTL on the two current mapping intervals but exclude the influence of QTL on other intervals and chromosomes. Epistatic QTL can be identified by ICIM, no matter whether the two interacting QTL have any additive effects. Simulated populations and one barley doubled haploids (DH) population were used to demonstrate the efficiency of ICIM in mapping both additive QTL and digenic interactions.     

An assessment of gene-by-gene interactions as a tool to unfold missing heritability in dyslexia

Even if substantial heritability has been reported and candidate genes have been identified extensively, all known marker associations explain only a small proportion of the phenotypic variance of developmental dyslexia (DD) and related quantitative phenotypes. Gene-by-gene interaction (also known as “epistasis”—G × G) triggers a non-additive effect of genes at different loci and should be taken into account in explaining part of the missing heritability of this complex trait. We assessed potential G × G interactions among five DD candidate genes, i.e., DYX1C1, DCDC2, KIAA0319, ROBO1, and GRIN2B, upon DD-related neuropsychological phenotypes in 493 nuclear families with DD, by implementing two complementary regression-based approaches: (1) a general linear model equation whereby the trait is predicted by the main effect of the number of rare alleles of the two genes and by the effect of the interaction between them, and (2) a family-based association test to detect G × G interactions between two unlinked markers by splitting up the association effect into a between- and a within-family genetic orthogonal components. After applying 500,000 permutations and correcting for multiple testing, both methods show that G × G effects between markers within the DYX1C1, KIAA0319/TTRAP, and GRIN2B genes lower the memory letters composite z-score of on average 0.55 standard deviation. We provided initial evidence that the effects of familial transmission of synergistic interactions between genetic risk variants can be exploited in the study of the etiology of DD, explain part of its missing heritability, and assist in designing customized charts of individualized neurocognitive impairments in complex disorders, such as DD.     

Conditional and unconditional QTL mapping of drought-tolerance-related traits of wheat seedling using two related RIL populations

For discovering the quantitative trait loci (QTLs) contributing to early seedling growth and drought tolerance during germination, conditional and unconditional analyses of 12 traits of wheat seedlings: coleoptile length, seedling height, longest root length, root number, seedling fresh weight, stem and leaves fresh weight, root fresh weight, seedling dry weight, stem and leaves dry weight, root dry weight, root to shoot fresh weight ratio, root-to-shoot dry weight ratio, were conducted under two water conditions using two F_8:9 recombinant inbred line (RIL) populations. The results of unconditional analysis are as follows: 88 QTLs accounting for 3.33–77.01% of the phenotypic variations were detected on chromosomes 1A, 1B, 1D, 2A, 2B, 2D, 3A, 3B, 4A, 4B, 4D, 5A, 5B, 5D, 6A, 6B, 6D, 7A, 7B and 7D. Among these QTLs, 19 were main-effect QTLs with a contribution rate greater than 10%. The results of the conditional QTL analysis of 12 traits under osmotic stress on normal water conditions were as follows: altogether 22 QTLs concerned with drought tolerance were detected on chromosomes 1B, 2A, 2B, 3B, 4A, 5D, 6A, 6D, 7B, and 7D. Of these QTLs, six were main-effect QTLs. These 22 QTLs were all special loci directly concerned with drought tolerance and most of them could not be detected by unconditional analysis. The finding of these QTLs has an important significance for fine-mapping technique, map-based cloning, and molecular marker-assisted selection of early seedling traits, such as growth and drought tolerance.     

A random model approach to mapping quantitative trait loci for complex binary traits in outbred populations.

Mapping quantitative trait loci (QTL) for complex binary traits is more challenging than for normally distributed traits due to the nonlinear relationship between the observed phenotype and unobservable genetic effects, especially when the mapping population contains multiple outbred families. Because the number of alleles of a QTL depends on the number of founders in an outbred population, it is more appropriate to treat the effect of each allele as a random variable so that a single variance rather than individual allelic effects is estimated and tested. Such a method is called the random model approach. In this study, we develop the random model approach of QTL mapping for binary traits in outbred populations. An EM-algorithm with a Fisher-scoring algorithm embedded in each E-step is adopted here to estimate the genetic variances. A simple Monte Carlo integration technique is used here to calculate the likelihood-ratio test statistic. For the first time we show that QTL of complex binary traits in an outbred population can be scanned along a chromosome for their positions, estimated for their explained variances, and tested for their statistical significance. Application of the method is illustrated using a set of simulated data.     

Estimates of heritability and correlations of morphometric traits in Clarkia (Onagraceae)

Summary Interspecific heritability values were estimated using parent-offspring regression analyses for 11 morphological traits differentiating Clarkia nitens and C. speciosa subsp. polyantha. Estimates ranged from near 0 for anther color and germination percentage, to 0.8 for calyx length and petal tip color. Phenotypic, genetic, and environmental correlation matrices were computed to determine the extent of interspecific correlations of traits. Cluster analyses of the genetic and environmental correlation matrices each resulted in three clusters of correlated traits; however, the clusters derived from the two matrices were different. The clusters produced by analysis of the environmental correlation matrix were similar to the factors obtained from principal component analysis of the phenotypic correlation matrix. Genetic correlations may result from strong linkage due to interspecific chromosomal differences.