The estimator of the optimal measure of allelic association: mean, variance and probability distribution when the sample size tends to infinity

The allelic association or linkage disequilibrium between two loci is a parameter of fundamental interest in modern population genetics for evolutionary inference and association mapping studies. Among the many measures available, the optimal measure of allelic association rho presents a strong evolutionary theory basis and is modeled on the physical distance along the chromosome with the Malécot equation for isolation by distance. Moreover, rho is equal to the absolute value of D', the standardized measure of gametic disequilibrium. We studied here the statistical properties of the rho sample estimator. We derived its asymptotic probability distribution and showed that it is neither asymptotically normal nor unbiased when rho=0 or when allelic frequencies are equal at both loci, in contrast to previous claims. This asymptotic study leads to propose a new test for absence of linkage disequilibrium. We compared it to Pearson's Chi2 test for independence in a contingency table and showed by simulations that the range in power of these two tests depends on the sign of D'. The new test outperformed slightly the Chi2 test, when D', polarized with respect to major alleles, is negative. Finally, we derived the asymptotic bias and information of the rho estimator that are due to the experimental sampling and showed by simulation that its bias is large in small samples. The consequences of these findings on applications using the rho measure are then discussed in particular for constructing LD unit maps, and call for a revised statistical treatment.     

Selection, Hitchhiking and Disequilibrium Analysis at Three Linked Loci with Application to Hla Data

The HLA system has been extensively studied from an evolutionary perspective. Although it is clear that selection has acted on the genes in the HLA complex, the nature of this selection has yet to be fully clarified. A study of constrained disequilibrium values is presented that is applicable to HLA and other less polymorphic systems with three or more linked loci, with the purpose of identifying selection events. The method uses the fact that three locus systems impose additional constraints on the range of possible disequilibrium values for any pair of loci. We have thus examined the behavior of the normalized pairwise disequilibrium measures using two locus (D'), and also three locus (D"), constraints on pairwise disequilibria in a three locus system when one of the three loci is under positive selection. The difference between these measures, delta = magnitude of D' - magnitude of D", has a distribution for the two unselected loci differing from that for the selected locus with either of the unselected loci (the hallmark is a high positive value of delta for the two unselected loci). An examination of genetic drift indicates that positive delta values are unlikely to be found in human populations in the absence of selection when recombination is greater than about 0.1%. This measure can thus provide insight into which allele of several linked loci might have been subject to selection. Application of this method to HLA haplotypes from a large French population study (Provinces Francaise) identifies selected alleles on particular haplotypes. Application of a complementary method, disequilibrium pattern analysis also confirms the action of selection on these haplotypes.     

On supergenes. II. The estimation of gametic excess in natural populations

The behaviour of supergenes (closely linked blocks of co-adapted loci) can be studied by using various parameters of gametic excess (= linkage disequilibrium). Three useful parameters are the gametic determinant (D), the relative determinant (D'), and the log of the eross product ratio. The behaviour of the gametic determinant under mating and recombination is described. Formulae are given for estimating gametic excess in natural populations, Recombination, and the fitness of double heterozygotes, may be estimated from the determinants at different stages of the life cycle; this would be useful in human population genetics.     

Inferences about linkage disequilibrium.

Existing theory for inferences about linkage disequilibrium is restricted to a measure defined on gametic frequencies. Unless gametic frequencies are directly observable, they are inferred from genotypic frequencies under the assumption of random union of gametes. Primary emphasis in this paper is given to genotypic data, and disequilibrium coefficients are defined for all subsets of two or more of the four genes, two at each of two loci, carried by an individual. Linkage disequilibrium coefficients are defined for genes within and between gametes, and methods of estimating and testing these coefficients are given for gametic data. For genotypic data, when coupling and repulsion double heterozygotes cannot be distinguished. Burrows' composite measure of linkage disequilibrium is discussed. In particular, the estimate for this measure and hypothesis tests based on it are compared to the usual maximum likelihood estimate of gametic linkage disequilibrium, and corresponding likelihood ratio or contingency chi-square tests. General use of the composite measure, whether or not random union of gametes is an appropriate assumption, is recommended. Attention is given to small samples, where the non-normality of gene frequencies will have greatest effect on methods of inference based on normal theory. Even tools such as Fisher's z-transformation for the correlation of gene frequencies are found to perform quite satisfactorily.     

A comparison of linkage disequilibrium patterns and estimated population recombination rates across multiple populations

Large-scale studies of linkage disequilibrium (LD) have shown considerable variation in the extent and distribution of pairwise LD within and between populations. Taken at face value, these results suggest that genomewide LD maps for one population may not be generalizable to other populations. However, at least part of this diversity is due to some undesirable features of pairwise LD measures, which are well documented for the D' and r2 measures. In this report, we compare patterns of LD derived from pairwise measures with statistical estimates of population recombination rates ( rho ) along a 10-Mb stretch of chromosome 20 in four population samples, comprising East Asians, African Americans, and U.K. and U.S. individuals of western European descent. The results reveal the expected variability of D' within and between populations but show better concordance in estimates of r2 for the same markers across the population samples. Estimates of rho correlate well across populations, but there is still evidence of population-specific spikes and troughs in rho values. We conclude that it is unlikely that a single haplotype map will provide a definitive guide for association studies of many populations; rather, multiple maps will need to be constructed to provide the best-possible guides for gene mapping.     

Homozygosity and linkage disequilibrium

We illustrate how homozygosity of haplotypes can be used to measure the level of disequilibrium between two or more markers. An excess of either homozygosity or heterozygosity signals a departure from the gametic phase equilibrium: We describe the specific form of dependence that is associated with high (low) homozygosity and derive various linkage disequilibrium measures. They feature a clear biological interpretation, can be used to construct tests, and are standardized to allow comparison across loci and populations. They are particularly advantageous to measure linkage disequilibrium between highly polymorphic markers.     

THE DETECTION OF GAMETIC DISEQUILIBRIUM BETWEEN ALLOZYME LOCI IN NATURAL POPULATIONS OF DROSOPHILA

The capacity of the usual tests (chi-square and related tests) to detect gametic disequilibrium between allozyme loci in natural populations of Drosophila has been investigated. We analyzed a large collection of previously reported gametic samples from natural populations involving a variety of loosely linked allozyme loci located along the O chromosome of Drosophila subobscura and the second chromosome of D. melanogaster. It is found that the statistical power of the individual tests to detect the sample disequilibria between allozyme loci is remarkably low, being the average (over pairs of loci) of power estimates close to 0.20 in both species. Moreover, the average minimum disequilibrium (D‘_min) that would be required to reject (90% probability) the hypothesis of gametic equilibrium is higher than 0.50 given the observed degree of polymorphism and sample sizes used. This means that statistically significant associations between allozyme loci would rarely be detected by single-sample tests even when much disequilibrium is present in natural populations of Drosophila. However, an alternative approach based on the analysis of disquilibrium for large sets of gametic samples, combining probabilities from single independent tests and assessing significance by a bootstrap procedure, reveals that most of the locus pairs within segment I and II of the O chromosome of D. subobscura and left arm of the second chromosome of D. melanogaster present significant nonrandom associations. Within these chromosomal sections, the observed average absolute value of disquilibrium (D‘) between loci is around 0.25 (under the more conservative estimation). Also, a positive relationship between the magnitude of disequilibrium and linkage was detected. These findings suggest that weak or moderate values of disequilibrium between loosely linked allozyme loci are more frequent in natural populations of Drosophila than is currently believed.     

Linkage disequilibrium and haplotype homozygosity in population samples genotyped at a high marker density

OBJECTIVE: Analyze the information contained in homozygous haplotypes detected with high density genotyping. METHODS: We analyze the genotypes of approximately 2,500 markers on chr 22 in 12 population samples, each including 200 individuals. We develop a measure of disequilibrium based on haplotype homozygosity and an algorithm to identify genomic segments characterized by non-random homozygosity (NRH), taking into account allele frequencies, missing data, genotyping error, and linkage disequilibrium. RESULTS: We show how our measure of linkage disequilibrium based on homozygosity leads to results comparable to those of R(2), as well as the importance of correcting for small sample variation when evaluating D'. We observe that the regions that harbor NRH segments tend to be consistent across populations, are gene rich, and are characterized by lower recombination. CONCLUSIONS: It is crucial to take into account LD patterns when interpreting long stretches of homozygous markers.     

The effect that genotyping errors have on the robustness of common linkage-disequilibrium measures

The rapid development of a dense single-nucleotide-polymorphism marker map has stimulated numerous studies attempting to characterize the magnitude and distribution of background linkage disequilibrium (LD) within and between human populations. Although genotyping errors are an inherent problem in all LD studies, there have been few systematic investigations documenting their consequences on estimates of background LD. Therefore, we derived simple deterministic formulas to investigate the effect that genotyping errors have on four commonly used LD measures-D', r, Q, and d-in studies of background LD. We have found that genotyping error rates as small as 3% can have serious affects on these LD measures, depending on the allele frequencies and the assumed error model. Furthermore, we compared the robustness of D', r, Q, and d, in the presence of genotyping errors. In general, Q and d are more robust than D' and r, although exceptions do exist. Finally, through stochastic simulations, we illustrate how genotyping errors can lead to erroneous inferences when measures of LD between two samples are compared.     

Genome-wide linkage disequilibrium in two Japanese beef cattle breeds

There is little knowledge about the degree of linkage disequilibrium (LD) in beef cattle. This study aims to perform a genome-wide search for LD in Japanese Black and Japanese Brown beef cattle and to compare the level of LD between these two breeds. Parameter D' (the LD coefficient) was used as a measure of LD, and LD was tested for significance of allelic associations between syntenic and between non-syntenic marker pairs. Effects of breed, chromosome, genetic map distance and their interactions with D' were tested based on least squares analyses. Both breeds showed high levels of LD, which ranged over several tens of cM and declined as the marker distance increased for syntenic marker pairs. A rapid decline of the D' value was observed between markers that were spaced 5 and 20 cM apart. LD was significant in most cases for marker pairs <40 cM apart but was not significant between non-syntenic loci. The pattern of LD found in these two breeds was similar to that previously published for dairy cattle. The D' value between breeds was not significantly different (P > 0.05), but the interaction between breed and chromosome was highly significant (P < 0.001). Genetic selection seems to have caused the heterogeneity of the D' values among chromosomes within breed. These results indicate that LD mapping is a useful tool for fine-mapping quantitative trait loci of economically important traits in Japanese beef cattle.     

Recombination and linkage disequilibrium among mitochondrial genes in structured populations of the gynodioecious plant Silene vulgaris

The impact of intergenic recombination on the population genetics of plant mitochondrial genomes is unknown. In an effort to study this in the gynodioecious plant Silene vulgaris three-locus PCR/RFLP genotypes (based on the mitochondrial genes atpA, cox1, and cob) were determined for 239 individuals collected from 20 North American populations. Seventeen three-locus PCR/RFLP genotypes were found. Recombination was indicated by observation of each of the four two-locus genotypes possible when the two most common alleles are considered for each of two loci. Based on these common alleles the absolute values of standardized linkage disequilibrium |D'| between pairs of loci range from 0.17 to 0.78. This indicates modest disequilibrium, rather than the maximum value expected in the absence of recombination |D'=1|, or the linkage equilibrium expected if recombination is pervasive (D'=0). Values of D' did not depend on which pair of loci contributed alleles to the analysis. The direction of D' obtained for the common atpA and cox1 alleles was comparable in sign and magnitude to that obtained by examining similar information obtained in a prior study of European samples. All three loci indicated a high degree of population structure (average FST=0.63), which would limit the within-population genetic diversity required for intergenic recombination to create novel genotypes, if most mating is local. Thus, population structure acts as a constraint on the approach to linkage equilibrium.     

Evaluation of linkage disequilibrium measures between multi-allelic markers as predictors of linkage disequilibrium between markers and QTL

Effectiveness of marker-assisted selection (MAS) and quantitative trait loci (QTL) mapping using population-wide linkage disequilibrium (LD) between markers and QTL depends on the extent of LD and how it declines with distance in a population. Because marker-QTL LD cannot be observed directly, the objective of this study was to evaluate alternative measures of observable LD between multi-allelic markers as predictors of usable LD of multi-allelic markers with presumed biallelic QTL. Observable LD between marker pairs was evaluated using eight existing measures and one new measure. These consisted of two pooled and standardized measures of LD between pairs of alleles at two markers based on Lewontin's LD measure, two pooled measures of squared correlations between alleles, one standardized measure using Hardy-Weinberg heterozygosities, and four measures based on the chi-square statistic for testing for association between alleles at two loci. In simulated populations with a range of LD generated by drift and a range of marker polymorphism, marker-marker LD measured by a standardized chi-square statistic (denoted chi(2')) was found to be the best predictor of useable marker-QTL LD for a group of multi-allelic markers. Estimates of the level and decline of marker-marker LD with distance obtained from chi(2') were linearly and highly correlated with usable LD of those markers with QTL across population structures and marker polymorphism. Corresponding relationships were poorer for the other marker-marker LD measures. Therefore, when LD is generated by drift, chi(2') is recommended to quantify the amount and extent of usable LD in a population for QTL mapping and MAS based on multi-allelic markers.     

Theoretical value of estimates of general combining ability in the autotetraploid crop

Summary The means of half-sib progenies have been indicated as selection criteria for intra-population improvement while the average of the means of full-sib progenies in diallel analyses have been proposed as predictors, in part, of the means of untested synthetic varieties. When these measures based on progeny means are expressed as deviations from a defined greater population of crosses, they are often termed the general combining ability (GCA). In this study the GCA estimates or a facsimile were theoretically investigated for the one locus, digene, autotetraploid model to verify the genetic basis and its value for selection and prediction in the presence of a naturally occurring phenomena of autopolyploids called gametic disequilibrium with three types of non-additive inheritance. Two breeding objectives were envisioned, the selection of best parents with recurrent selection based on GCA in the continued development of elite populations and the prediction of advanced generation synthetic variety performance. The first generation means of progenies with a potential bias due to gametic disequilibrium were compared to GCA estimation of same progenies in the absence of gametic disequilibrium. The results indicated that testcrossing plants to a population without gametic disequilibrium could be used for selection of best parents. The gametic disequilibrium in the cross may increase or depress selection response dependent on the array of genotypes which happen to be evaluated, on the type of genic action at the locus, and on the frequency of the desirable allele in the testor population. The GCA estimates for prediction of synthetic performance were potentially biased by gametic disequilibrium. An assumption of pollination by the same array of gametes was made for all plants, but obviously was unrealistic for GCA estimation with partial diallels, or with no selfing, and in other situations. The GCA estimate was shown to be an unreliable predictor of synthetic variety performance. When it was assumed that different plants were pollinated by different arrays of gametes, a more realistic situation, no genetic interpretation of GCA values was possible even with purely additive gene action at the locus.Cooperative investigation of the Alfalfa Production Research Unit, United States Department of Agriculture, Agricultural Research Service, and the Nevada Agricultural Experiment Station, Reno, Nevada     

Association studies in consanguineous populations.

To study the genetic determinism of multifactorial diseases in large panmictic populations, a strategy consists in looking for an association with markers closely linked to candidate genes. A distribution of marker genotypes different in patients and controls may indicate that the candidate gene is involved in the disease. In panmictic populations, the power to detect the role of a candidate gene depends on the gametic disequilibrium with the marker locus. In consanguineous populations, we show that it depends on the inbreeding coefficient F as well. Inbreeding increases the power to detect the role of a recessive or quasi-recessive disease-susceptibility factor. The gain in power turns out to be greater for small values of the gametic disequilibrium. Moreover, even in the absence of gametic disequilibrium, the presence of inbreeding may allow to detect the role of a recessive factor. Ignoring inbreeding when it exists may lead to reject falsely a recessive model if the mode of inheritance is inferred on the distribution of genotypes among patients.     

On the detection of nonrandom associations between DNA polymorphisms in natural populations of Drosophila

The capacity to detect nonrandom associations between restriction-map variants was examined in eight gene regions of Drosophila melanogaster (yellow-achaetescute, white, Zw, Adh, Est6, and rosy) and D. pseudoobscura (Adh and Xdh), on the basis of published population data. The statistical power from individual pairwise tests was both heterogeneous and generally low across gene regions. Sample sizes larger than those currently being used are needed to ensure any power to detect disequilibrium by individual tests. It is found that the heterogeneity in power is mostly explained by large differences in the intensity of sample disequilibrium among regions. The yellow-achaete- scute, Zw, and Adh loci of D. melanogaster displayed both the highest mean power (approximately 0.4) and a very great disequilibrium (mean absolute values of D' were 0.8-1). By contrast, all the other gene regions exhibited lower mean power (approximately 0.2) and moderate levels of disequilibrium (0.4-0.6). Although the proportion of significant pairwise associations, especially for white, Est6, and rosy in D. melanogaster and for Adh and Xdh in D. pseudoobscura, is more or less close to the type I error, simultaneous-inference significance tests show that gametic disequilibrium is occurring at the eight DNA regions examined.      

Linkage disequilibrium in natural populations of Trypanosoma cruzi (flagellate), the agent of Chagas' disease

We have studied linkage disequilibrium in natural populations of Trypanosoma cruzi, the agent of Chagas' disease, by analyzing (i) a set of 524 stocks from the whole geographical range of the parasite, characterized at four gene loci coding for enzymes; (ii) a subsample of 121 stocks characterized at 12 enzyme loci; and (iii) a subset of 386 stocks from six locations in Bolivia, characterized by four enzyme loci. Our results show that the linkage disequilibrium reaches the maximum possible value, given the observed allelic frequencies, for almost all the locus pairs. This result is most consistent with the hypothesis that genetic recombination is absent or very rare in T. cruzi natural populations. Partition of the linkage disequilibrium variance for the six Bolivian populations shows that both inter- and intrapopulation components are substantial and that the relationships among the components are D2IS less than D2ST, and D'2IS less than D'2ST. These inequalities are interpreted as the result of an interplay between genetic drift, rare or absent mating, and clonal selection in generating linkage disequilibrium in T. cruzi populations.     

Genetic microstructure in two Spanish cat populations. II: gametic disequilibrium and spatial autocorrelation

In a previous publication, we described some aspects of the microgenetic structure of two Spanish cat populations (in Barcelona and Alicante). In the present study, the possible existence ofgametic disequilibrium and spatial genetic structure for these populations, at the coat colour pattern and length genes O, A, T D, L, S and W, was analyzed. There was little gametic disequilibrium between pairs of these loci, despite certain pairs that showed significant systematic gametic disequilibrium (a-d and O-S), which appears to show the action of natural selection on domestic cat populations. Nevertheless, we believe that the major cause of the small amount of gametic disequilibrium found was probably a combination of gene drift and gene flow. The results obtained here were clearly in disagreement with those of Hedrick (1985), who concluded that epistatic selection was the cause of the gametic disequilibrium that he found in cat populations. We also found that although Hardy-Weinberg equilibrium could not be demonstrated, the gametic disequilibrium statistics were not affected by this fact, adding credence to the estimates obtained. We found no genetic spatial structure inside the city of Barcelona, as shown by analysis of the spatial autocorrelation of the individual loci, and analysis of the coordinates of the two first axes of a multidimensional scale. However, some gametic disequilibrium statistics showed certain spatial patterns, which leads us to consider the possibility of several evolutionary processes acting upon some of Barcelona's cat colonies.     

Multilocus and Multitrait Measures of Differentiation for Gene Markers and Phenotypic Traits

Multilocus measures of differentiation taking into account gametic disequilibrium are developed. Even if coupling and repulsion heterozygotes cannot be separated at the multilocus level, a method is given to calculate a composite measure of differentiation (CF(st)) at the zygotic level, which accounts for allelic associations combining both gametic and nongametic effects. Mean and maximum differentiations may be relevant when multilocus measures are computed. Maximum differentiation is the highest eigenvalue of the F(st) matrix, whereas mean differentiation corresponds to the mean value of all eigenvalues of the F(st) matrix. Gametic disequilibrium has a stronger effect on maximum differentiation than on mean differentiation and takes into account the anisotropy that may exist between within- and between-population components of disequilibria. Multilocus mean and maximum differentiation are calculated for a set of 81 Quercus petraea (sessile oak) populations assessed with eight allozyme loci and two phenotypic traits (bud burst and height growth). The results indicate that maximum differentiation increases as more loci (traits) are considered whereas mean differentiation remains constant or decreases. Phenotypic traits exhibit higher population differentiation than allozymes. The applications and uses of mean and maximum differentiations are further discussed.     

Comparison of estimators for measures of linkage disequilibrium

The measurement of biallelic pair-wise association called linkage disequilibrium (LD) is an important issue in order to understand the genomic architecture. A plethora of measures of association in two by two tables have been proposed in the literature. Beside the problem of choosing an appropriate measure, the problem of their estimation has been neglected in the literature. It needs to be emphasized that the definition of a measure and the choice of an estimator function for it are conceptually unrelated tasks. In this paper, we compare the performance of various estimators for the three popular LD measures D', r and Y in a simulation study for small to moderate samples sizes (N<=500). The usual frequency-plug-in estimators can lead to unreliable or undefined estimates. Estimators based on the computationally expensive volume measures have been proposed recently as a remedy to this well-known problem. We confirm that volume estimators have better expected mean square error than the naive plug-in estimators. But they are outperformed by estimators plugging-in easy to calculate non-informative Bayesian probability estimates into the theoretical formulae for the measures. Fully Bayesian estimators with non-informative Dirichlet priors have comparable accuracy but are computationally more expensive. We recommend the use of non-informative Bayesian plug-in estimators based on Jeffreys' prior, in particular when dealing with SNP array data where the occurrence of small table entries and table margins is likely.     

Definition and Properties of Disequilibrium Statistics for Associations between Nuclear and Cytoplasmic Genotypes

We define and establish the interrelationships of four components of statistical association between a diploid nuclear gene and a uniparentally transmitted, haploid cytoplasmic gene: an allelic (gametic) disequilibrium (D), which measures associations between alleles at the two loci; and three genotypic disequilibria (D1, D2, D3), which measure associations between two cytotypes and the three respective nuclear backgrounds. We also consider an alternative set of measures, including D and the residual disequilibrium (d). The dynamics of these disequilibria are then examined under three conventional models of the mating system: (1) random mating; (2a) assortative mating without dominance (the "mixed-mating model"); and (2b) assortative mating with dominance ("O'Donald's model"). The trajectories of gametic disequilibria are similar to those for pairs of unlinked nuclear loci. The dynamics of genotypic disequilibria exhibit a variety of behaviors depending on the model and the initial conditions. Procedures for statistical estimation of cytonuclear disequilibria are developed and applied to several real and hypothetical data sets. Special attention is paid to the biological interpretations of various categories of allelic and genotypic disequilibria in hybrid zones. Genetic systems for which these statistics might be appropriate include nuclear genotype frequencies in conjunction with those for mitochondrial DNA, chloroplast DNA, or cytoplasmically inherited microorganisms.