Variance components models for gene-environment interaction in quantitative trait locus linkage analysis.

BackgroundWe analyzed the Genetic Analysis Workshop 13 (GAW13) simulated data to contrast and compare different methods for the genetic linkage analysis of hypertension and change in blood pressure over time. We also examined methods for incorporating covariates into the linkage analysis. We used methods for quantitative trait loci (QTL) linkage analysis with and without covariates and affected sib-pair (ASP) analysis of hypertension followed by ordered subset analysis (OSA), using variables associated with change in blood pressure over time.ResultsFour of the five baseline genes and one of the three slope genes were not detected by any method using conventional criteria. OSA detected baseline gene b35 on chromosome 13 when using the slope in blood pressure to adjust for change over time. Slope gene s10 was detected by the ASP analysis and slope gene s11 was detected by QTL linkage analysis as well as by OSA analysis. Analysis of null chromosomes, i.e., chromosomes without genes, did not reveal significant increases in type I error. However, there were a number of genes indirectly related to blood pressure detected by a variety of methods.ConclusionsWe noted that there is no obvious first choice of analysis software for analyzing a complicated model, such as the one underlying the GAW13 simulated data. Inclusion of covariates and longitudinal data can improve localization of genes for complex traits but it is not always clear how best to do this. It remains a worthwhile task to apply several different approaches since one method is not always the best.

Variance components modeling has emerged as a powerful method for quantitative trait loci (QTL) linkage analysis. However, the power to detect a gene of minor effect is low. Many complex traits are affected by environmental as well as genetic factors, and one strategy to increase power is to reduce nongenetic phenotypic variance by adjusting for environmental covariates. In this paper, we investigate the power of three approaches to covariate adjustment in variance components linkage analysis: (i) incorporate covariates in the means model, (ii) incorporate covariates in the covariance matrix, and (iii) perform analysis on residual statistics. These approaches are compared to an analysis without adjustment. The results show that in the absence of correlation between the covariate and the QTL effect, adjusting for covariates indeed increases power to detect an underlying QTL. As this correlation increases, however, the power decreases. In the presence of a causal association between QTL and covariates, not adjusting for covariates appeared to be more powerful. The three approaches for adjusting for covariate: residual statistics, the means model, and the covariance model, had equal power to detect a QTL.

Both lipoprotein-associated phospholipase A(2) (Lp-PLA(2)) activity, a biomarker of inflammation, and concentration of its primary associated lipoprotein, LDL, are correlated with adverse coronary outcomes. We previously reported a quantitative trait locus (QTL) corresponding to HSA2p24.3-p23.2 with pleiotropic effects on Lp-PLA(2) activity and LDL-cholesterol (LDL-C) concentration in baboons fed a basal diet. Here, our goal was to locate pleiotropic QTLs influencing both traits in the same baboons fed a high-cholesterol, high-fat (HCHF) diet, and to assess whether shared genetic effects on these traits differ between diets. We assayed Lp-PLA(2) activity and LDL-C concentration in 683 baboons fed the HCHF diet. We used a bivariate maximum likelihood-based variance components approach in whole-genome linkage screens to locate a QTL [logarithm of odds (LOD) = 3.13, genome-wide P = 0.019] corresponding to HSA19q12-q13.2 with pleiotropic effects on Lp-PLA(2) activity and LDL-C levels in the HCHF diet. We additionally found significant evidence of genetic variance in response to diet for Lp-PLA(2) activity (P = 0.0017) and for LDL-C concentration (P = 0.00001), revealing a contribution of genotype-by-diet interaction to covariation in these two traits. We conclude that the pleiotropic QTLs detected at 2p24.3-p23.2 and 19q12-q13.2 on the basal and HCHF diets, respectively, exert diet-specific effects on covariation in Lp-PLA(2) activity and LDL-C concentration.

BackgroundVariance component (VC) models are commonly used for Quantitative Trait Loci (QTL) mapping in outbred populations. Here, the QTL effect is given as a random effect and a critical part of the model is the relationship between the phenotypic values and the random effect. In the traditional VC model, each individual has a unique QTL effect and the relationship between these random effects is given as a covariance structure (known as the identity-by-descent (IBD) matrix).ResultsWe present an alternative notation of the variance component model, where the elements of the random effect are independent base generation allele effects and sampling term effects. The relationship between the phenotypic vales and the random effect is given by an incidence matrix, which results in a novel, but statistically equivalent, version of the traditional VC model. A general algorithm to estimate this incidence matrix is presented. Since the model is given in terms of base generation allele effects and sampling term effects, these effects can be estimated separately using best linear unbiased prediction (BLUP). From simulated data, we showed that biallelic QTL effects could be accurately clustered using the BLUP obtained from our model notation when markers are fully informative, and that the accuracy increased with the size of the QTL effect. We also developed a measure indicating whether a base generation marker homozygote is a QTL heterozygote or not, by comparing the variances of the sampling term BLUP and the base generation allele BLUP. A ratio greater than one gives strong support for a QTL heterozygote.ConclusionWe developed a simple presentation of the VC QTL model for identification of base generation allele effects in QTL linkage analysis. The base generation allele effects and sampling term effects were separated in our model notation. This clarifies the assumptions of the model and should also enhance the development of genome scan methods.

To increase the likelihood of finding genetic variation conferring liability to eating disorders, we measured over 100 attributes thought to be related to liability to eating disorders on affected individuals from multiplex families and two cohorts: one recruited through a proband with anorexia nervosa (AN; AN cohort); the other recruited through a proband with bulimia nervosa (BN; BN cohort). By a multilayer decision process based on expert evaluation and statistical analysis, six traits were selected for linkage analysis (1): obsessionality (OBS), age at menarche (MENAR), and anxiety (ANX) for quantitative trait locus (QTL) linkage analysis; and lifetime minimum body mass index (BMI), concern over mistakes (CM), and food-related obsessions (OBF) for covariate-based linkage analysis. The BN cohort produced the largest linkage signals: for QTL linkage analysis, four suggestive signals: (for MENAR, at 10p13; for ANX, at 1q31.1, 4q35.2, and 8q13.1); for covariate-based linkage analyses, both significant and suggestive linkages (for BMI, one significant [4q21.1] and three suggestive [3p23, 10p13, 5p15.3]; for CM, two significant [16p13.3, 14q21.1] and three suggestive [4p15.33, 8q11.23, 10p11.21]; and for OBF, one significant [14q21.1] and five suggestive [4p16.1, 10p13.1, 8q11.23, 16p13.3, 18p11.31]). Results from the AN cohort were far less compelling: for QTL linkage analysis, two suggestive signals (for OBS at 6q21 and for ANX at 9p21.3); for covariate-based linkage analysis, five suggestive signals (for BMI at 4q13.1, for CM at 11p11.2 and 17q25.1, and for OBF at 17q25.1 and 15q26.2). Overlap between the two cohorts was minimal for substantial linkage signals.

Key messageUsing QTL analysis and fine mapping, the novel recessive gene xa44(t) conferring resistance to BB was identified and the expression level of the gene was confirmed through qRT-PCR analysis.Bacterial blight (BB) disease caused by Xanthomonas oryzae pv. oryzae (Xoo) is a major factor causing rice yield loss in most rice-cultivating countries, especially in Asia. The deployment of cultivars with resistance to BB is the most effective method to control the disease. However, the evolution of new Xoo or pathotypes altered by single-gene-dependent mutations often results in breakdown of resistance. Thus, efforts to identify novel R-genes with sustainable BB resistance are urgently needed. In this study, we identified three quantitative trait loci (QTLs) on chromosomes 1, 4, and 11, from an F2 population of 493 individuals derived from a cross between IR73571-3B-11-3-K3 and Ilpum using a 7K SNP chip. Of these QTLs, one major QTL, qBB_11, on chromosome 11 explained 61.58% of the total phenotypic variance in the population, with an LOD value of 113.59, based on SNPs 11964077 and 11985463. The single major R-gene, with recessive gene action, was designated xa44(t) and was narrowed down to a 120-kb segment flanked within 28.00 Mbp to 28.12 Mbp. Of nine ORFs present in the target region, two ORFs revealed significantly different expression levels of the candidate genes. These candidate genes (Os11g0690066 and Os11g0690466) are described as “serine/threonine protein kinase domain containing protein” and “hypothetical protein,” respectively. The results will be useful to further understand BB resistance mechanisms and provide new sources of resistance, together with DNA markers for MAS breeding to improve BB resistance in rice.

One of the significant factors that affect brain dopamine function is the activity of tyrosine hydroxylase (TH), the first and rate-limiting enzyme in catecholamine biosynthesis. For the analysis of the genetically determined role of dopamine function and TH in behavior and in the regulatory mechanisms of the mesotelencephalic dopamine system we devised a novel genetic strategy (Vadasz; Mouse Genome 88:16-18; 1990). We hypothesized that phenotypic introgression and recombinant fixation could ensure the transfer of Quantitative Trait Loci (QTL) from one strain onto the genetic background of another strain, and new, genetically very similar quasi-congenic strains could be created that would carry individual QTLs, or QTLs in various combinations. Here we summarize the construction of the first set of QTL Introgression strains, and present evidence that QTLs that are responsible for the continuous variation of mesencephalic tyrosine hydroxylase activity (TH/MES), have been transferred onto the C57BL/6By (B6) strain background from BALB/cJ (C) and CXBI (I) donor strains with high and low TH/MES, respectively. The QTL transfer was carried out in two directions by repeated backcross-intercross cycles with concomitant selection for the extreme high and low expressions of TH/MES in replicates, resulting in four QTL Introgression lines. Analysis of regional brain TH activities in the course of the QTL introgression indicated that (a) TH activity in B6.I lines showed quite limited heritability, (b) TH/MES was not highly correlated with striatal TH, and (c) the control of hypothalamic and olfactory tubercle TH activities was largely independent from that of TH/MES. Examination of the open-field (OF) behavior data demonstrated that TH activity did not correlate significantly with OF behavior. After 5 backcross-intercross cycles, TH/MES in each replicate line was still significantly different from that of the B6 background strain. A genomewide scanning of microsatellite markers in the QTL introgression lines demonstrated that about 96% of the markers were of background (B6) type. These results indicate the successful transfer of TH/MES QTLs. After the QTL transfer phase of the experiment altogether more than 100 new RQI strains were initiated in the QTL Introgression lines by strict brother x sister mating. After fixing the introgressed QTLs, ten of the inbred RQI strains were tested for TH/MES. The results showed that in one of the new RQI strains TH/MES was restored to a level that is characteristic to the C donor strain, while TH/MES values in some other strains were between those of the background and donor strains, confirming our hypothesis that phenotypic introgression and recombinant fixation can ensure a virtually complete transfer of QTLs. We conclude from this study that complex, continuously distributed neural traits can successfully be subjected to QTL introgression, and the results raise the possibility that the RQI method can be efficiently applied for gene mapping of complex neural and behavioral traits even if their phenotypic expression is sensitive to confounding developmental and environmental variations, genetic interactions, and genotype-environment interactions.

The common assumption in quantitative trait locus (QTL) linkage mapping studies that parents of multiple connected populations are unrelated is unrealistic for many plant breeding programs. We remove this assumption and propose a Bayesian approach that clusters the alleles of the parents of the current mapping populations from locus-specific identity by descent (IBD) matrices that capture ancestral marker and pedigree information. Moreover, we demonstrate how the parental IBD data can be incorporated into a QTL linkage analysis framework by using two approaches: a Threshold IBD model (TIBD) and a Latent Ancestral Allele Model (LAAM). The TIBD and LAAM models are empirically tested via numerical simulation based on the structure of a commercial maize breeding program. The simulations included a pilot dataset with closely linked QTL on a single linkage group and 100 replicated datasets with five linkage groups harboring four unlinked QTL. The simulation results show that including parental IBD data (similarly for TIBD and LAAM) significantly improves the power and particularly accuracy of QTL mapping, e.g., position, effect size and individuals’ genotype probability without significantly increasing computational demand.

Advances in cereal genomics and applications in crop breeding

Abstract. The main objective was to locate QTL and estimate the proportion of total genetic variance attributable to quantitative trait loci (QTL) for production index traits and the udder health index identified on six Bos taurus autosomes in the Danish Holstein dairy cattle population. Data were obtained from a granddaughter design of 20 sire families with a total of 1 869 progeny tested sons. The number of sons per grandsire ranged from 20 to 284, with an average family size of 93.5. Indexes of the estimated breeding values were obtained for the milk production traits and for the udder health index from the Danish Agricultural Advisory Service database. A random-QTL model was applied to incorporate marker information into parameter estimation for each single QTL. The procedure allowed us to detect new QTL on BTA3, BTA16 and BTA28 and to estimate the proportion of total genetic variance attributed to different QTL on a total of six Bos taurus autosomes for the udder health index and yield index traits in the Danish Holstein population. Variance estimates vary between 2 to 58 % of the total variance for different QTL and seem to explain a substantial part of the variance at certain positions of the cattle genome. The results are discussed against the background of the failure of marker-assisted selection (MAS) and the recent availability of large panels of single nucleotide polymorphisms (SNPs) that have improved the search for mutations underlying variation in complex traits resulting in modern genomic selection.

Trudy Mackay

Using a novel approach for a two-locus model that provides a greatly increased power to detect multiple quantitative trait loci (QTLs) in simulated data, we identified in a sample of 961 female sib-pairs, three genome-wide significant QTLs for apolipoprotein A1 on chromosomes 8p21.1-q13.1 (LOD score 3.71), 9q21.32-33.1 (LOD score 3.28) and 10p15.1-p13 (LOD score 5.51), two for lipoprotein (a) on chromosomes 6q25.2-q27 (LOD score 10.18) and 21q21.1-q21.3 (LOD score 4.57) and two for triglycerides on chromosomes 4q28.3-32.1 (LOD score 3.71) and 5q23.1-q32 (LOD score 3.60). The two-locus ordered-subset analysis has led to the confirmation of known and likely identification of novel regions linked to serum lipid levels that would have otherwise been missed and deserves wider application in linkage analyses of quantitative traits. Given the relative lack of power for the sample sizes commonly used in human genetics linkage studies, minor QTL effects often go undetected and those that are detected will be upwardly biased. We show through simulation that the discrepancy between the real and estimated QTL-effects is often likely to generate an unpredictable source of false-negative errors, using multi-locus models, reducing the power to detect multiple QTLs through oligogenic linkage analysis. The successful simultaneous modelling of the identified QTLs in a multi-locus context helps to eliminate false positives and increases the power to detect linkages, adding compelling evidence that they are likely to be reliable QTLs for these lipid traits.

The genetic basis of variation in complex traits remains poorly understood, and few genes underlying variation have been identified. Previous work identified a quantitative trait locus (QTL) responsible for much of the response to selection on growth in mice, effecting a change in body mass of approximately 20%. By fine-mapping, we have resolved the location of this QTL to a 660-kb region containing only two genes of known function, Gpc3 and Gpc4, and two other putative genes of unknown function. There are no non-synonymous polymorphisms in any of these genes, indicating that the QTL affects gene regulation. Mice carrying the high-growth QTL allele have approximately 15% lower Gpc3 mRNA expression in kidney and liver, whereas expression differences at Gpc4 are non-significant. Expression profiles of the two other genes within the region are inconsistent with a factor responsible for a general effect on growth. Polymorphisms in the 3′ untranslated region of Gpc3 are strong candidates for the causal sequence variation. Gpc3 loss-of-function mutations in humans and mice cause overgrowth and developmental abnormalities. However, no deleterious side-effects were detected in our mice, indicating that genes involved in Mendelian diseases also contribute to complex trait variation. Furthermore, these findings show that small changes in gene expression can have substantial phenotypic effects.

QTLs with epistatic effects and environmental interaction effects for the developmental behavior of plant height in rice were studied by conventional and conditional methods for quantitative trait loci (QTLs) by mapping with a doubled-haploid population of 123 lines from IR64/Azucena in three environments. The results showed that epistatic effects were important and most epistasis could be detected only by conditional QTL mapping, while most non-epistatic QTLs could be de- tected by both conventional and conditional methods. Many modificative QTLs showed only epistatic effects without their own additive effects at some stages. QTL× environment (QE) interaction effects were detect- ed more often than QTL main effects for plant-height be- havior, which might indicate that gene expression could be greatly affected by the environment. No QTLs had ef- fects during the whole of ontogeny. Conditional QTL mapping might be a valid way to reveal dynamic gene expression for the development of quantitative traits, es- pecially for epistatic effects.

Strategies for mapping quantitative trait (QT) loci (QTL) need special studies in species where inbred lines cannot easily be obtained. We propose the detection of QTL with schemes including several full-sib families and study here the efficiency of such schemes — diallel, factorial, cyclic or single-pair mating designs — for a varying number of parents, on the basis of the power of tests for QTL detection. We considered a single, fully informative marker completely linked to the QTL. The tests proposed for QTL detection are based on linear additive models. The results show that for a given number of parents, in given conditions, the mating design has no effect on power. It increases as the total number of individuals genotyped increases. The number of parents affects the power of QTL detection insofar as the effect of substituting QT alleles varies from one parent to another and as some parents can be homozygous at the QTL: with only two parents QTL detection is often less powerful than with more parents. The size of the genetic effects that can be detected with a given power decreases as the number of parents increases from two to six and is quite stable for more than six parents. If QTL heterozygote frequency in the base population is high enough (greater than 0.2) mating designs with six parents should give a good sample of variance attributable to QTL and allow the detection of QTL with reasonable power.