Haplotype Heritability Mapping Method Uncovers Missing Heritability of Complex Traits.

We propose a novel approach to analyze genomic data that incorporates haplotype information for detecting rare variants within a regional heritability mapping framework. The performance of our approach was tested in a simulation study based on human genotypes. The phenotypes were simulated by generating regional variance using either SNP(s) or haplotype(s). Regional genomic relationship matrices, constructed with either a SNP-based or a haplotype-based estimator, were employed to estimate the regional variance. The results from the study show that haplotype heritability mapping captures the regional effect, with its relative performance decreasing with increasing analysis window size. The SNP-based regional mapping approach often misses the effect of causal haplotype(s); however, it has a greater power to detect simulated SNP-based-variants. Heritability estimates suggest that the haplotype heritability mapping estimates the simulated regional heritability accurately for all phenotypes and analysis windows. However, the SNP-based analysis overestimates the regional heritability and performs less well than our haplotype-based approach for the simulated rare haplotype-based-variant. We conclude that haplotype heritability mapping is a useful tool to capture the effect of rare variants, and explain a proportion of the missing heritability.

Comparative analysis of quantitative trait loci for body weight, growth rate and growth curve parameters from 3 to 72 weeks of age in female chickens of a broiler-layer cross.

BACKGROUND: Comparisons of quantitative trait loci (QTL) for growth and parameters of growth curves assist in understanding the genetics and ultimately the physiology of growth. Records of body weight at 3, 6, 12, 24, 48 and 72 weeks of age and growth rate between successive age intervals of about 500 F2 female chickens of the Roslin broiler-layer cross were available for analysis. These data were analysed to detect and compare QTL for body weight, growth rate and parameters of the Gompertz growth function. RESULTS: Over 50 QTL were identified for body weight at specific ages and most were also detected in the nearest preceding and/or subsequent growth stage. The sum of the significant and suggestive additive effects for bodyweight at specific ages accounted for 23-43% of the phenotypic variation. A single QTL for body weight on chromosome 4 at 48 weeks of age had the largest additive effect (550.4 ± 68.0 g, 11.5% of the phenotypic variation) and a QTL at a similar position accounted 14.5% of the phenotypic variation at 12 weeks of age. Age specific QTL for growth rate were detected suggesting that there are specific genes that affect developmental processes during the different stages of growth. Relatively few QTL influencing Gompertz growth curve parameters were detected and overlapped with loci affecting growth rate. Dominance effects were generally not significant but from 12 weeks of age they exceeded the additive effect in a few cases. No evidence for epistatic QTL pairs was found. CONCLUSIONS: The results confirm the location for body weight and body weight gain during growth that were identified in previous studies and were consistent with QTL for the parameters of the Gompertz growth function. Chromosome 4 explained a relatively large proportion of the observed growth variation across the different ages, and also harboured most of the detected QTL for Gompertz parameters, confirming its importance in controlling growth. Very few QTL were detected for body weight or gain at 48 and 72 weeks of age, probably reflecting the effect of differences in reproduction and random environmental effects.

Genome-wide analysis of epistasis in body mass index using multiple human populations.

We surveyed gene-gene interactions (epistasis) in human body mass index (BMI) in four European populations (n<1200) via exhaustive pair-wise genome scans where interactions were computed as F ratios by testing a linear regression model fitting two single-nucleotide polymorphisms (SNPs) with interactions against the one without. Before the association tests, BMI was corrected for sex and age, normalised and adjusted for relatedness. Neither single SNPs nor SNP interactions were genome-wide significant in either cohort based on the consensus threshold (P=5.0E-08) and a Bonferroni corrected threshold (P=1.1E-12), respectively. Next we compared sub genome-wide significant SNP interactions (P<5.0E-08) across cohorts to identify common epistatic signals, where SNPs were annotated to genes to test for gene ontology (GO) enrichment. Among the epistatic genes contributing to the commonly enriched GO terms, 19 were shared across study cohorts of which 15 are previously published genome-wide association loci, including CDH13 (cadherin 13) associated with height and SORCS2 (sortilin-related VPS10 domain containing receptor 2) associated with circulating insulin-like growth factor 1 and binding protein 3. Interactions between the 19 shared epistatic genes and those involving BMI candidate loci (P<5.0E-08) were tested across cohorts and found eight replicated at the SNP level (P<0.05) in at least one cohort, which were further tested and showed limited replication in a separate European population (n>5000). We conclude that genome-wide analysis of epistasis in multiple populations is an effective approach to provide new insights into the genetic regulation of BMI but requires additional efforts to confirm the findings.

Characterisation of genome-wide association epistasis signals for serum uric acid in human population isolates.

Genome-wide association (GWA) studies have identified a number of loci underlying variation in human serum uric acid (SUA) levels with the SLC2A9 gene having the largest effect identified so far. Gene-gene interactions (epistasis) are largely unexplored in these GWA studies. We performed a full pair-wise genome scan in the Italian MICROS population (n = 1201) to characterise epistasis signals in SUA levels. In the resultant epistasis profile, no SNP pairs reached the Bonferroni adjusted threshold for the pair-wise genome-wide significance. However, SLC2A9 was found interacting with multiple loci across the genome, with NFIA-SLC2A9 and SLC2A9-ESRRAP2 being significant based on a threshold derived for interactions between GWA significant SNPs and the genome and jointly explaining 8.0% of the phenotypic variance in SUA levels (3.4% by interaction components). Epistasis signal replication in a CROATIAN population (n = 1772) was limited at the SNP level but improved dramatically at the gene ontology level. In addition, gene ontology terms enriched by the epistasis signals in each population support links between SUA levels and neurological disorders. We conclude that GWA epistasis analysis is useful despite relatively low power in small isolated populations.

Quantitative trait loci for variation in immune response to a Foot-and-Mouth Disease virus peptide.

BACKGROUND: Infectious disease of livestock continues to be a cause of substantial economic loss and has adverse welfare consequences in both the developing and developed world. New solutions to control disease are needed and research focused on the genetic loci determining variation in immune-related traits has the potential to deliver solutions. However, identifying selectable markers and the causal genes involved in disease resistance and vaccine response is not straightforward. The aims of this study were to locate regions of the bovine genome that control the immune response post immunisation. 195 F2 and backcross Holstein Charolais cattle were immunised with a 40-mer peptide derived from foot-and-mouth disease virus (FMDV). T cell and antibody (IgG1 and IgG2) responses were measured at several time points post immunisation. All experimental animals (F0, F1 and F2, n = 982) were genotyped with 165 microsatellite markers for the genome scan. RESULTS: Considerable variability in the immune responses across time was observed and sire, dam and age had significant effects on responses at specific time points. There were significant correlations within traits across time, and between IgG1 and IgG2 traits, also some weak correlations were detected between T cell and IgG2 responses. The whole genome scan detected 77 quantitative trait loci (QTL), on 22 chromosomes, including clusters of QTL on BTA 4, 5, 6, 20, 23 and 25. Two QTL reached 5% genome wide significance (on BTA 6 and 24) and one on BTA 20 reached 1% genome wide significance. CONCLUSIONS: A proportion of the variance in the T cell and antibody response post immunisation with an FDMV peptide has a genetic component. Even though the antigen was relatively simple, the humoral and cell mediated responses were clearly under complex genetic control, with the majority of QTL located outside the MHC locus. The results suggest that there may be specific genes or loci that impact on variation in both the primary and secondary immune responses, whereas other loci may be specifically important for early or later phases of the immune response. Future fine mapping of the QTL clusters identified has the potential to reveal the causal variations underlying the variation in immune response observed.

Genetic comparison of a Croatian isolate and CEPH European founders.

Human isolates have been postulated as a good resource for the identification of QTL due to reduced genetic diversity and a more homogeneous environment. Isolates may also have increased linkage disequilibrium (LD) due to small effective population size and, either loss or increase in frequency of alleles that are rare in the general population from which they originate. Here we investigate the difference in allele and genotype frequencies, LD and homozygous tracts between an isolate-several villages from the island of Vis in Croatia-and an outbred population of European origin: the Hapmap CEPH founders. Using the HumanHap300 v1 Genotyping BeadChip, we show that our population does not differ greatly from the reference CEU outbred population despite having a slightly higher proportion of monomorphic loci, a slightly higher long-range LD, and a greater proportion of individuals with long homozygous tracts. We conclude that genotyping arrays should perform equally well in our isolate as in outbred European populations for disease mapping studies and that SNP-trait associations discovered in our well-characterized Croatian isolate should be valid in the general European population from which they descend.

Detecting parent of origin and dominant QTL in a two-generation commercial poultry pedigree using variance component methodology.

INTRODUCTION: Variance component QTL methodology was used to analyse three candidate regions on chicken chromosomes 1, 4 and 5 for dominant and parent-of-origin QTL effects. Data were available for bodyweight and conformation score measured at 40 days from a two-generation commercial broiler dam line. One hundred dams were nested in 46 sires with phenotypes and genotypes on 2708 offspring. Linear models were constructed to simultaneously estimate fixed, polygenic and QTL effects. Different genetic models were compared using likelihood ratio test statistics derived from the comparison of full with reduced or null models. Empirical thresholds were derived by permutation analysis. RESULTS: Dominant QTL were found for bodyweight on chicken chromosome 4 and for bodyweight and conformation score on chicken chromosome 5. Suggestive evidence for a maternally expressed QTL for bodyweight and conformation score was found on chromosome 1 in a region corresponding to orthologous imprinted regions in the human and mouse. CONCLUSION: Initial results suggest that variance component analysis can be applied within commercial populations for the direct detection of segregating dominant and parent of origin effects.

Detecting dominant QTL with variance component analysis in simulated pedigrees.

Dominance is an important source of variation in complex traits. Here, we have carried out the first thorough investigation of quantitative trait locus (QTL) detection using variance component (VC) models extended to incorporate both additive and dominant QTL effects. Simulation results showed that the empirical distribution of the test statistic when testing for dominant QTL effects did not behave in accordance with existing theoretical expectations and varied with pedigree structure. Extensive simulations were carried out to assess accuracy of estimates, type 1 error and statistical power in two-generation human-, poultry- and pig-type pedigrees each with 1900 progeny in small-, medium- and large-sized families, respectively. The distribution of the likelihood-ratio test statistic was heavily dependent on family structure, with empirical thresholds lower for human pedigrees. Power to detect QTL was high (0.84-1.0) in pig and poultry scenarios for dominance effects accounting for >7% of phenotypic variance but much lower (0.42) in human-type pedigrees. Maternal or common environment effects can be partially confounded with dominance and must be fitted in the QTL model. Including dominance in the QTL model did not affect power to detect additive QTL effects. Also, detection of spurious dominance QTL effects only occurred when maternal effects were not included in the QTL model. When dominance effects were present in the data but not in the analysis model, this resulted in spurious detection of additive QTL or inflated estimates of additive QTL effects. The study demonstrates that dominance can be included routinely in QTL analysis of general pedigrees; however, optimal power is dependent on selection of the appropriate thresholds for pedigree structure.

Genetic variation affecting host-parasite interactions: major-effect quantitative trait loci affect the transmission of sigma virus in Drosophila melanogaster.

In natural populations, genetic variation affects resistance to disease. Whether that genetic variation comprises lots of small-effect polymorphisms or a small number of large-effect polymorphisms has implications for adaptation, selection and how genetic variation is maintained in populations. Furthermore, how much genetic variation there is, and the genes that underlie this variation, affects models of co-evolution between parasites and their hosts. We are studying the genetic variation that affects the resistance of Drosophila melanogaster to its natural pathogen--the vertically transmitted sigma virus. We have carried out three separate quantitative trait locus mapping analyses to map gene variants on the second chromosome that cause variation in the rate at which males transmit the infection to their offspring. All three crosses identified a locus in a similar chromosomal location that causes a large drop in the rate at which the virus is transmitted. We also found evidence for an additional smaller-effect quantitative trait locus elsewhere on the chromosome. Our data, together with previous experiments on the sigma virus and parasitoid wasps, indicate that the resistance of D. melanogaster to co-evolved pathogens is controlled by a limited number of major-effect polymorphisms.

SLC2A9 is a newly identified urate transporter influencing serum urate concentration, urate excretion and gout.

Uric acid is the end product of purine metabolism in humans and great apes, which have lost hepatic uricase activity, leading to uniquely high serum uric acid concentrations (200-500 microM) compared with other mammals (3-120 microM). About 70% of daily urate disposal occurs via the kidneys, and in 5-25% of the human population, impaired renal excretion leads to hyperuricemia. About 10% of people with hyperuricemia develop gout, an inflammatory arthritis that results from deposition of monosodium urate crystals in the joint. We have identified genetic variants within a transporter gene, SLC2A9, that explain 1.7-5.3% of the variance in serum uric acid concentrations, following a genome-wide association scan in a Croatian population sample. SLC2A9 variants were also associated with low fractional excretion of uric acid and/or gout in UK, Croatian and German population samples. SLC2A9 is a known fructose transporter, and we now show that it has strong uric acid transport activity in Xenopus laevis oocytes.

Multiple-trait QTL mapping for body and organ weights in a cross between NMRI8 and DBA/2 mice.

Multiple-trait analyses have been shown to improve the detection of quantitative trait loci (QTLs) with multiple effects. Here we applied a multiple-trait approach on obesity- and growth-related traits that were surveyed in 275 F2 mice generated from an intercross between the high body weight selected line NMRI8 and DBA/2 as lean control. The parental lines differed 2.5-fold in body weight at the age of 6 weeks. Within the F2 population, the correlations between body weight and weights of abdominal fat weight, muscle, liver and kidney at the age of 6 weeks were about 0.8. A least squares multiple-trait QTL analysis was performed on these data to understand more precisely the cause of the genetic correlation between body weight, body composition traits and weights of inner organs. Regions on Chr 1, 2, 7 and 14 for body weights at different early ages and regions on Chr 1, 2, 4, 7, 14, 17 and 19 for organ weights at 6 weeks were found to have significant multiple effects at the genome-wide level.

False disease region identification from identity-by-descent haplotype sharing in the presence of phenocopies.

Linkage analysis (either parametric or nonparametric) is commonly applied to identify chromosomal regions using related individuals affected by disease. In complex disease the incomplete relationship between phenotype and genotype can be modeled using a phenocopy parameter, the probability that an individual is affected given they do not carry the disease mutation of interest, and a nonpenetrance parameter, the probability that an individual is not affected given they do carry the disease mutation of interest. If the linkage phase between multiple markers and a putative disease locus is known, then haplotypes carrying the mutation can, in principle, be identified by comparing the chromosome segments that are shared identical-by-descent (IBD) across affected individuals. We consider here the effect of a nonzero phenocopy rate on the linkage peak and hence upon the identification of disease haplotypes that are shared IBD between affected individuals. We show, by theory and computer simulation, that in diseases for which there is a nonzero phenocopy rate, the chromosomal regions identified may not include the true disease locus. We utilize a LOD-1 confidence interval for a widely used nonparametric linkage statistic. We find that in small/moderate samples this confidence interval may be inappropriate. We give specific examples where the phenocopy rates are nonnegligible in some complex diseases. The success of further work to identify the causal mutations underlying the linkage peaks in these diseases will depend on researchers allowing for the presence of phenocopies by examining appropriately wide regions around the initial positive linkage finding.

Quantitative trait locus analysis of longitudinal quantitative trait data in complex pedigrees.

There is currently considerable interest in genetic analysis of quantitative traits such as blood pressure and body mass index. Despite the fact that these traits change throughout life they are commonly analyzed only at a single time point. The genetic basis of such traits can be better understood by collecting and effectively analyzing longitudinal data. Analyses of these data are complicated by the need to incorporate information from complex pedigree structures and genetic markers. We propose conducting longitudinal quantitative trait locus (QTL) analyses on such data sets by using a flexible random regression estimation technique. The relationship between genetic effects at different ages is efficiently modeled using covariance functions (CFs). Using simulated data we show that the change in genetic effects over time can be well characterized using CFs and that including parameters to model the change in effect with age can provide substantial increases in power to detect QTL compared with repeated measure or univariate techniques. The asymptotic distributions of the methods used are investigated and methods for overcoming the practical difficulties in fitting CFs are discussed. The CF-based techniques should allow efficient multivariate analyses of many data sets in human and natural population genetics.

Regression-based quantitative trait loci mapping: robust, efficient and effective.

Regression has always been an important tool for quantitative geneticists. The use of maximum likelihood (ML) has been advocated for the detection of quantitative trait loci (QTL) through linkage with molecular markers, and this approach can be very effective. However, linear regression models have also been proposed which perform similarly to ML, while retaining the many beneficial features of regression and, hence, can be more tractable and versatile than ML in some circumstances. Here, the use of linear regression to detect QTL in structured outbred populations is reviewed and its perceived shortfalls are revisited. It is argued that the approach is valuable now and will remain so in the future.

Mapping quantitative trait loci using linkage disequilibrium: marker- versus trait-based methods.

Two approaches for mapping quantitative trait loci (QTL) using linkage disequilibrium at the population level were investigated. In the trait-based (TB) approach, the frequencies of marker alleles (or genotypes) are compared in individuals selected from the two tails of the trait distribution. The TB approach uses phenotypic information only in the selection step. In the marker-based (MB) approach, the quantitative trait values for the marker genotypes in the selected individuals are compared. The MB approach uses both the difference in marker allele (or genotype) frequencies and the phenotypic values of each marker genotype in the selected samples. We quantify the power of each approach and show that the power of the MB approach is greater than or equal to that of the TB approach. The advantage of the former is expected to increase with increasing number of traits phenotyped. Our accurate predictions obviate the need for elaborate simulation studies.

Identification and reciprocal introgression of a QTL affecting body mass in mice.

The aim of this study was to examine the effects of a QTL in different genetic backgrounds. A QTL affecting body mass on chromosome 6 was identified in an F(2) cross between two lines of mice that have been divergently selected for this trait. The effect of the QTL on mass increased between 6 and 10 weeks of age and was not sex-specific. Body composition analysis showed effects on fat-free dry body mass and fat mass. To examine the effect of this QTL in different genetic backgrounds, the high body mass sixth chromosome was introgressed into the low body mass genetic background and vice versa by repeated marker-assisted backcrossing. After three generations of backcrossing, new F(2) populations were established within each of the introgression lines by crossing individuals that were heterozygous across the sixth chromosome. The estimated additive effect of the QTL on 10-week body mass was similar in both genetic backgrounds and in the original F(2) population (i.e., approximately 0.4 phenotypic standard deviations); no evidence of epistatic interaction with the genetic background was found. The 95% confidence interval for the location of the QTL was refined to a region of approximately 7 cM between D6Mit268 and D6Mit123.

Theoretical and empirical power of regression and maximum-likelihood methods to map quantitative trait loci in general pedigrees.

Both theoretical calculations and simulation studies have been used to compare and contrast the statistical power of methods for mapping quantitative trait loci (QTLs) in simple and complex pedigrees. A widely used approach in such studies is to derive or simulate the expected mean test statistic under the alternative hypothesis of a segregating QTL and to equate a larger mean test statistic with larger power. In the present study, we show that, even when the test statistic under the null hypothesis of no linkage follows a known asymptotic distribution (the standard being chi(2)), it cannot be assumed that the distribution under the alternative hypothesis is noncentral chi(2). Hence, mean test statistics cannot be used to indicate power differences, and a comparison between methods that are based on simulated average test statistics may lead to the wrong conclusion. We illustrate this important finding, through simulations and analytical derivations, for a recently proposed new regression method for the analysis of general pedigrees to map quantitative trait loci. We show that this regression method is not necessarily more powerful nor computationally more efficient than a maximum-likelihood variance-component approach. We advocate the use of empirical power to compare trait-mapping methods.

Longitudinal variance-components analysis of the Framingham Heart Study data.

The Framingham Heart Study offspring cohort, a complex data set with irregularly spaced longitudinal phenotype data, was made available as part of Genetic Analysis Workshop 13. To allow an analysis of all of the data simultaneously, a mixed-model- based random-regression (RR) approach was used. The RR accounted for the variation in genetic effects (including marker-specific quantitative trait locus (QTL) effects) across time by fitting polynomials of age. The use of a mixed model allowed both fixed (such as sex) and random (such as familial environment) effects to be accounted for appropriately. Using this method we performed a QTL analysis of all of the available adult phenotype data (26,106 phenotypic records). In addition to RR, conventional univariate variance component techniques were applied. The traits of interest were BMI, HDLC, total cholesterol, and height. The longitudinal method allowed the characterization of the change in QTL effects with aging. A QTL affecting BMI was shown to act mainly at early ages.

QTL Express: mapping quantitative trait loci in simple and complex pedigrees.

QTL Express is the first application for Quantitative Trait Locus (QTL) mapping in outbred populations with a web-based user interface. User input of three files containing a marker map, trait data and marker genotypes allows mapping of single or multiple QTL by the regression approach, with the option to perform permutation or bootstrap tests.