Strategies for genetic mapping of categorical traits

The search for efficient and powerful statistical methods and optimal mapping strategies for categorical traits under various experimental designs continues to be one of the main tasks in genetic mapping studies. Methodologies for genetic mapping of categorical traits can generally be classified into two groups, linear and non-linear models. We develop a method based on a threshold model, termed mixture threshold model to handle ordinal (or binary) data from multiple families. Monte Carlo simulations are done to compare its statistical efficiencies and properties of the proposed non-linear model with a linear model for genetic mapping of categorical traits using multiple families. The mixture threshold model has notably higher statistical power than linear models. There may be an optimal sampling strategy (family size vs number of families) in which genetic mapping reaches its maximal power and minimal estimation errors. A single large-sibship family does not necessarily produce the maximal power for detection of quantitative trait loci (QTL) due to genetic sampling of QTL alleles. The QTL allelic model has a marked impact on efficiency of genetic mapping of categorical traits in terms of statistical power and QTL parameter estimation. Compared with a fixed number of QTL alleles (two or four), the model with an infinite number of QTL alleles and normally distributed allelic effects results in loss of statistical power. The results imply that inbred designs (e.g. F₂ or four-way crosses) with a few QTL alleles segregating or reducing number of QTL alleles (e.g. by selection) in outbred populations are desirable in genetic mapping of categorical traits using data from multiple families. This revised version was published online in July 2006 with corrections to the Cover Date.     

The role of maternal and paternal effects in the evolution of parental quality by sexual selection

Genetic models of maternal effects and models of mate choice have focused on the evolutionary effects of variation in parental quality. There have been, however, few attempts to combine these into a single model for the evolution of sexually selected traits. We present a quantitative genetic model that considers how male and female parental quality (together or separately) affect the expression of a sexually selected offspring trait. We allow female choice of males based on this parentally affected trait and examine the evolution of mate choice, parental quality and the indicator trait. Our model reveals a number of consequences of maternal and paternal effects. (1) The force of sexual selection owing to adaptive mate choice can displace parental quality from its natural selection optimum. (2) The force of sexual selection can displace female parental quality from its natural selection optimum even when nonadaptive mate choice occurs (e.g. runaway sexual selection), because females of higher parental quality produce more attractive sons and these sons counterbalance the loss in fitness owing to over-investment in each offspring. (3) Maternal and paternal effects can provide a source of genetic variation for offspring traits, allowing evolution by sexual selection even when those traits do not show direct genetic variation (i.e. are not heritable). (4) The correlation between paternal investment and the offspring trait influenced by the parental effects can result in adaptive mate choice and lead to the elaboration of both female preference and the male sexually selected trait. When parental effects exist, sexual selection can drive the evolution of parental quality when investment increases the attractiveness of offspring, leading to the elaboration of indicator traits and higher than expected levels of parental investment.     

Recent developments in statistical methods for detecting genetic loci affecting phenotypic variability

ABSTRACT: A number of recent works have introduced statistical methods for detecting genetic loci that affect phenotypic variability, which we refer to as variability-controlling quantitative trait loci (vQTL). These are genetic variants whose allelic state predicts how much phenotype values will vary about their expected means. Such loci are of great potential interest in both human and non-human genetic studies, one reason being that a detected vQTL could represent a previously undetected interaction with other genes or environmental factors. The simultaneous publication of these new methods in different journals has in many cases precluded opportunity for comparison. We survey some of these methods, the respective trade-offs they imply, and the connections between them. The methods fall into three main groups: classical non-parametric, fully parametric, and semi-parametric two-stage approximations. Choosing between alternatives involves balancing the need for robustness, flexibility, and speed. For each method, we identify important assumptions and limitations, including those of practical importance, such as their scope for including covariates and random effects. We show in simulations that both parametric methods and their semi-parametric approximations can give elevated false positive rates when they ignore mean-variance relationships intrinsic to the data generation process. We conclude that choice of method depends on the trait distribution, the need to include non-genetic covariates, and the population size and structure, coupled with a critical evaluation of how these fit with the assumptions of the statistical model.     

A Multi-Marker Model for Detecting Chromosomal Segments Displaying Qtl Activity

A statistical method is presented for detecting quantitative trait loci (QTLs), based on the linear model. Unlike methods able to detect a few well separated QTLs and to estimate their effects and positions, this method considers the genome as a whole and enables the detection of chromosomal segments involved in the differences between two homozygous lines, and their backcross, doubled haploid, or F(2) progenies, for a quantitative trait. Genetic markers must be codominant, but missing markers are accepted, provided they are missing independently from the experiment. Asymptotic properties, which are of practical use, are developed. This method does not rely on strong genetic hypotheses, and thus does not permit any precise genetic analysis of the trait under study, but it does assess which regions of the genome are involved, whatever the complexity of the genetic determinism (number, effects and interactions among QTLs). Simultaneous use of several methods, including this one, should lead to better efficiency in QTL detection.     

The evolution of indicator traits for parental quality: the role of maternal and paternal effects

In systems where individuals provide material resources to their mates or offspring, mate choice based on traits that are phenotypically correlated with the quality of resources provided is expected to be adaptive. Several models have explored the evolution of mating preference where there are direct benefits to choice, but few have addressed how a phenotypic correlation can be established between a male indicator trait and the degree of parental investment. We present a model with three quantitative traits: male and female parental investment and a potential male indicator trait. In our model, the expression of the "indicator" trait in offspring is affected by parental investment. These effects are referred to as maternal or paternal effects, or as "indirect genetic effects" when parental investment is heritable. With genetic variation for degree of parental investment, offspring harbor genes for parental investment that are unexpressed before mating but will affect the investment that they provide when expressed. Because the investment received from the parents affects the expression of the indicator trait, there will be a correlation between the genes for parental investment inherited and the degree of expression of the indicator trait in the offspring. The indicator trait is thus an "honest" signal for the degree of paternal investment.     

By-product runaway evolution by adaptive mate choice: A behavioural aspect of sexual selection

Summary From a behavioural perspective on adaptive female choice, I developed a by-product runaway model of adaptive mate choice. The model illustrates the evolution of the tail size of peacocks. I consider the causal mechanisms of adaptive female choice: (1) why (ultimate reasons); (2) how (proximate mechanisms). Assumptions are developed based on these behavioural aspects. For (1) ultimate reasons, I assume that many male losers (low-fitness males) always occur due to genetic and environmental uncertainty (A-1). For (2) proximate mechanisms, I assume that losers tend to differ in the expression of a fitness-sensitive trait (an ultimate target, e.g. body size; A-2), that the fitness-sensitive trait correlates with a secondary sexual trait (a proximate cue, e.g. allometry in body size and tail size; (A-3), and that the cue trait has a genetic basis that is independent of the target trait (e.g. a genetic basis in tail ratio to body size; A-4). The model's results are: persistent female choice by means of a proximate cue (R-1); by-product selection on the independent genetic basis of the cue (R-2); and the non-adaptive or maladaptive runaway evolution of the male proximate cue (R-3). In this model, female mate preferences are non-arbitrary and adaptive, whereas the resulting evolution of male secondary sexual traits is non-adaptive in the sense of survival selection.     

Robust multipoint identical-by-descent mapping for affected relative pairs

The genetic mapping of complex traits has been challenging and has required new statistical methods that are robust to misspecified models. Liang et al. proposed a robust multipoint method that can be used to simultaneously estimate, on the basis of sib-pair linkage data, both the position of a trait locus on a chromosome and its effect on disease status. The advantage of their method is that it does not require specification of an underlying genetic model, so estimation of the position of a trait locus on a specified chromosome and of its standard error is robust to a wide variety of genetic mechanisms. If multiple loci influence the trait, the method models the marginal effect of a locus on a specified chromosome. The main critical assumption is that there is only one trait locus on the chromosome of interest. We extend this method to different types of affected relative pairs (ARPs) by two approaches. One approach is to estimate the position of a trait locus yet allow unconstrained trait-locus effects across different types of ARPs. This robust approach allows for differences in sharing alleles identical-by-descent across different types of ARPs. Some examples for which an unconstrained model would apply are differences due to secular changes in diagnostic methods that can change the frequency of phenocopies among different types of relative pairs, environmental factors that modify the genetic effect, epistasis, and variation in marker-information content. However, this unconstrained model requires a parameter for each type of relative pair. To reduce the number of parameters, we propose a second approach that models the marginal effect of a susceptibility locus. This constrained model is robust for a trait caused by either a single locus or by multiple loci without epistasis. To evaluate the adequacy of the constrained model, we developed a robust score statistic. These methods are applied to a prostate cancer-linkage study, which emphasizes their potential advantages and limitations.     

GEOGRAPHIC VARIATION IN FEMALE PREFERENCES FOR MALE TRAITS IN POECILIA RETICULATA

We examined the preferences of female guppies (Poecilia reticulata) from 11 localities in Trinidad with respect to male color-pattern elements, body shape and size, and overall color and brightness contrast. Females are on average more attracted to males from their own population than from alien populations, and populations appear to vary in the criteria used in female choice. Multiple-regression analysis suggests that mate-preference criteria vary among localities in intensity, sign, and the number of traits used. Although preference estimators and color-pattern parameters are unique to each population, only orange, black, and color contrast showed a correlation between degree of male trait and degree of preference for that trait. There is a clear effect of water color and a possible effect of predation intensity. The results are discussed in light of various models of sexual selection and the early stages of speciation.     

Patterns of fluctuating asymmetry in sexual ornaments predict female choice

Extravagant secondary sexual characters are assumed to have arisen and be maintained by sexual selection. While traits like horns, antlers and spurs can be ascribed to intrasexual competition, other traits such as extravagant feather ornaments, displays and pheromones have to be ascribed to mate choice. A number of studies have tested whether females exert selection on the size of male ornaments, but only some of these have recorded female preferences for the most extravagantly ornamented males. Here I demonstrate that female choice can be directly predicted from the relationship between the degree of fluctuating asymmetry and the size of a secondary sexual character. Fluctuating asymmetry is an epigenetic measure of the ability of individuals to cope with stress, and it occurs when an individual is unable to undergo identical development of an otherwise bilaterally symmetric trait on both sides of its body. There is a negative relationship between the degree of fluctuating asymmetry and the absolute size of an ornament in those bird species with a female preference for the largest male sex trait, while there is a flat or U-shaped relationship among species without a female preference. These results suggest that females prefer exaggerated secondary sexual characters if they reliably demonstrate the ability of males to cope with genetic and environmental stress. Some species may demonstrate a flat or U-shaped relationship between the degree of fluctuating asymmetry and the absolute size of an ornament because (i) the genetic variance in viability signalled by the secondary sex trait has been depleted; (ii) the secondary sex trait is not particularly costly and therefore does not demonstrate condition dependence; or because (iii) the sex traits can be considered arbitrary traits rather than characters reflecting good genes.     

The evolution of phenotypes and genetic parameters under preferential mating

This article extends and adds more realism to Lande's analytical model for evolution under mate choice by using individual‐based simulations in which females sample a finite number of males and the genetic architecture of the preference and preferred trait evolves. The simulations show that the equilibrium heritabilities of the preference and preferred trait and the genetic correlation between them (r_G), depend critically on aspects of the mating system (the preference function, mode of mate choice, choosiness, and number of potential mates sampled), the presence or absence of natural selection on the preferred trait, and the initial genetic parameters. Under some parameter combinations, preferential mating increased the heritability of the preferred trait, providing a possible resolution for the lek paradox. The Kirkpatrick–Barton approximation for r_G proved to be biased downward, but the realized genetic correlations were also low, generally <0.2. Such low values of r_G indicate that coevolution of the preference and preferred trait is likely to be very slow and subject to significant stochastic variation. Lande's model accurately predicted the incidence of runaway selection in the simulations, except where preferences were relative and the preferred trait was subject to natural selection. In these cases, runaways were over‐ or underestimated, depending on the number of males sampled. We conclude that rapid coevolution of preferences and preferred traits is unlikely in natural populations, but that the parameter combinations most conducive to it are most likely to occur in lekking species.     

A statistical model for testing the pleiotropic control of phenotypic plasticity for a count trait

The differences of a phenotypic trait produced by a genotype in response to changes in the environment are referred to as phenotypic plasticity. Despite its importance in the maintenance of genetic diversity via genotype-by-environment interactions, little is known about the detailed genetic architecture of this phenomenon, thus limiting our ability to predict the pattern and process of microevolutionary responses to changing environments. In this article, we develop a statistical model for mapping quantitative trait loci (QTL) that control the phenotypic plasticity of a complex trait through differentiated expressions of pleiotropic QTL in different environments. In particular, our model focuses on count traits that represent an important aspect of biological systems, controlled by a network of multiple genes and environmental factors. The model was derived within a multivariate mixture model framework in which QTL genotype-specific mixture components are modeled by a multivariate Poisson distribution for a count trait expressed in multiple clonal replicates. A two-stage hierarchic EM algorithm is implemented to obtain the maximum-likelihood estimates of the Poisson parameters that specify environment-specific genetic effects of a QTL and residual errors. By approximating the number of sylleptic branches on the main stems of poplar hybrids by a Poisson distribution, the new model was applied to map QTL that contribute to the phenotypic plasticity of a count trait. The statistical behavior of the model and its utilization were investigated through simulation studies that mimic the poplar example used. This model will provide insights into how genomes and environments interact to determine the phenotypes of complex count traits.     

The effects of intraspecific competition and stabilizing selection on a polygenic trait

The equilibrium properties of an additive multilocus model of a quantitative trait under frequency- and density-dependent selection are investigated. Two opposing evolutionary forces are assumed to act: (i) stabilizing selection on the trait, which favors genotypes with an intermediate phenotype, and (ii) intraspecific competition mediated by that trait, which favors genotypes whose effect on the trait deviates most from that of the prevailing genotypes. Accordingly, fitnesses of genotypes have a frequency-independent component describing stabilizing selection and a frequency- and density-dependent component modeling competition. We study how the equilibrium structure, in particular, number, degree of polymorphism, and genetic variance of stable equilibria, is affected by the strength of frequency dependence, and what role the number of loci, the amount of recombination, and the demographic parameters play. To this end, we employ a statistical and numerical approach, complemented by analytical results, and explore how the equilibrium properties averaged over a large number of genetic systems with a given number of loci and average amount of recombination depend on the ecological and demographic parameters. We identify two parameter regions with a transitory region in between, in which the equilibrium properties of genetic systems are distinctively different. These regions depend on the strength of frequency dependence relative to pure stabilizing selection and on the demographic parameters, but not on the number of loci or the amount of recombination. We further study the shape of the fitness function observed at equilibrium and the extent to which the dynamics in this model are adaptive, and we present examples of equilibrium distributions of genotypic values under strong frequency dependence. Consequences for the maintenance of genetic variation, the detection of disruptive selection, and models of sympatric speciation are discussed.     

A Pleiotropic Nonadditive Model of Variation in Quantitative Traits

A model of mutation-selection-drift balance incorporating pleiotropic and dominance effects of new mutations on quantitative traits and fitness is investigated and used to predict the amount and nature of genetic variation maintained in segregating populations. The model is based on recent information on the joint distribution of mutant effects on bristle traits and fitness in Drosophila melanogaster from experiments on the accumulation of spontaneous and P element-induced mutations. These experiments suggest a leptokurtic distribution of effects with an intermediate correlation between effects on the trait and fitness. Mutants of large effect tend to be partially recessive while those with smaller effect are on average additive, but apparently with very variable gene action. The model is parameterized with two different sets of information derived from P element insertion and spontaneous mutation data, though the latter are not fully known. They differ in the number of mutations per generation which is assumed to affect the trait. Predictions of the variance maintained for bristle number assuming parameters derived from effects of P element insertions, in which the proportion of mutations with an effect on the trait is small, fit reasonably well with experimental observations. The equilibrium genetic variance is nearly independent of the degree of dominance of new mutations. Heritabilities of between 0.4 and 0.6 are predicted with population sizes from 10(4) to 10(6), and most of the variance for the metric trait in segregating populations is due to a small proportion of mutations (about 1% of the total number) with neutral or nearly neutral effects on fitness and intermediate effects on the trait (0.1-0.5σ(P)). Much of the genetic variance is contributed by recessive or partially recessive mutants, but only a small proportion (about 10%) of the genetic variance is dominance variance. The amount of apparent selection on the trait itself generated by the model is very small. If a model is assumed in which all mutation events have an effect on the quantitative trait, the majority of the genetic variance is contributed by deleterious mutations with tiny effects on the trait. If such a model is assumed for viability, the heritability is about 0.1, independent of the population size.     

Unraveling the actual background of second litter syndrome in pigs: based on Large White data

Second litter syndrome (SLS) in sows is when fertility performance is lower in the second parity than in the first parity. The causes of SLS have been associated with lactation weight loss, premature first insemination, short lactation length, short weaning to insemination interval, season, and farm of farrowing. There is little known about the genetic background of SLS or if it is a real biological problem or just a statistical issue. Thus, we aimed to evaluate risk factors, investigate genetic background of SLS, and estimate the probability of SLS existing due to the statistical properties of the trait. The records of 246 799 litters (total number born, TNB) from 46 218 Large White sows were used. A total of 15 398 sows had SLS. Two traits were defined: first a binominal trait if a sow had SLS or not (biSLS) and second a continuous trait (Range) created by subtracting the total number of piglets born in the first parity (TNB1) from the piglets born in the second parity (TNB2). Lactation length, farm, and season of the farrowing had significant effects on SLS traits when tested as fixed effects in the genetic model. These effects are farm management-related factors. The age at first insemination and weaning to insemination interval were significant only for other reproduction traits (e.g., TNB1, TNB2, litter weight in parity 1 and 2). The heritability of biSLS was 0.05 (on observed scale), whereas heritability of Range was 0.03. To verify the existence of SLS data with records of 50 000 sows and 9 parities was simulated. The simulations showed that the average expected frequency of SLS across all the parities was 0.49 (±0.05) while the observed frequency in the actual data was 0.46 (±0.04). We compared this to SLS frequencies in 67 farms and only 2 farms had more piglets born in the first parity compared to the second. Therefore, on the individual sow level SLS is likely due to statistical properties of the trait, whereas on the farm level SLS is likely due to farm management. Thus, SLS should not be considered an abnormality nor a syndrome if on average the herd litter size in parity 2 is larger than in parity 1.     

The costs of choice in sexual selection

In Fisher's model of sexual selection female mating preferences are not subject to direct selection but evolve purely because they are genetically correlated with the favoured male trait. But when female choice is costly relative to random mating, for example in energy, time or predation risks, the evolution of female mating preference is subject also to direct selection. With costly female choice the set or line of equilibria found in models of Fisher's process no longer exists. On the line the male trait is under zero net selection, and there is no advantage for a female choosing a male with a more exaggerated character. Therefore any cost to choice causes choosiness to decline. In turn this lowers the strength of sexual selection and the male trait declines as well. So when Fisher's process is the sole force of sexual selection and female choice is costly, only transitory increases in female choice and the preferred male trait are possible. It has often been claimed that exaggerated male characters act as markers or revealers of the genetic quality of potential mates. If females choose their mates using traits that correlate with heritable viability differences then stable exaggeration of both female choice and the preferred male character is possible, even when female choice is costly. The offspring of choosy females have not only a Fisherian reproductive advantage but also greater viability. This suggests that in species with exaggerated male ornamentation, in which female choice is costly, it is likely that female mate choice will be for traits that correlate with male genetic quality.     

Quantitative trait nucleotide analysis using Bayesian model selection

Although much attention has been given to statistical genetic methods for the initial localization and fine mapping of quantitative trait loci (QTLs), little methodological work has been done to date on the problem of statistically identifying the most likely functional polymorphisms using sequence data. In this paper we provide a general statistical genetic framework, called Bayesian quantitative trait nucleotide (BQTN) analysis, for assessing the likely functional status of genetic variants. The approach requires the initial enumeration of all genetic variants in a set of resequenced individuals. These polymorphisms are then typed in a large number of individuals (potentially in families), and marker variation is related to quantitative phenotypic variation using Bayesian model selection and averaging. For each sequence variant a posterior probability of effect is obtained and can be used to prioritize additional molecular functional experiments. An example of this quantitative nucleotide analysis is provided using the GAW12 simulated data. The results show that the BQTN method may be useful for choosing the most likely functional variants within a gene (or set of genes). We also include instructions on how to use our computer program, SOLAR, for association analysis and BQTN analysis.     

A dynamic model for genome-wide association studies

Although genome-wide association studies (GWAS) are widely used to identify the genetic and environmental etiology of a trait, several key issues related to their statistical power and biological relevance have remained unexplored. Here, we describe a novel statistical approach, called functional GWAS or fGWAS, to analyze the genetic control of traits by integrating biological principles of trait formation into the GWAS framework through mathematical and statistical bridges. fGWAS can address many fundamental questions, such as the patterns of genetic control over development, the duration of genetic effects, as well as what causes developmental trajectories to change or stop changing. In statistics, fGWAS displays increased power for gene detection by capitalizing on cumulative phenotypic variation in a longitudinal trait over time and increased robustness for manipulating sparse longitudinal data.     

A conceptual framework for mapping quantitative trait Loci regulating ontogenetic allometry

Although ontogenetic changes in body shape and its associated allometry has been studied for over a century, essentially nothing is known about their underlying genetic and developmental mechanisms. One of the reasons for this ignorance is the unavailability of a conceptual framework to formulate the experimental design for data collection and statistical models for data analyses. We developed a framework model for unraveling the genetic machinery for ontogenetic changes of allometry. The model incorporates the mathematical aspects of ontogenetic growth and allometry into a maximum likelihood framework for quantitative trait locus (QTL) mapping. As a quantitative platform, the model allows for the testing of a number of biologically meaningful hypotheses to explore the pleiotropic basis of the QTL that regulate ontogeny and allometry. Simulation studies and real data analysis of a live example in soybean have been performed to investigate the statistical behavior of the model and validate its practical utilization. The statistical model proposed will help to study the genetic architecture of complex phenotypes and, therefore, gain better insights into the mechanistic regulation for developmental patterns and processes in organisms.     

Pleiotropic Models of Polygenic Variation, Stabilizing Selection, and Epistasis

We show that in polymorphic populations many polygenic traits pleiotropically related to fitness are expected to be under apparent ``stabilizing selection' independently of the real selection acting on the population. This occurs, for example, if the genetic system is at a stable polymorphic equilibrium determined by selection and the nonadditive contributions of the loci to the trait value either are absent, or are random and independent of those to fitness. Stabilizing selection is also observed if the polygenic system is at an equilibrium determined by a balance between selection and mutation (or migration) when both additive and nonadditive contributions of the loci to the trait value are random and independent of those to fitness. We also compare different viability models that can maintain genetic variability at many loci with respect to their ability to account for the strong stabilizing selection on an additive trait. Let V(m) be the genetic variance supplied by mutation (or migration) each generation, V(g) be the genotypic variance maintained in the population, and n be the number of the loci influencing fitness. We demonstrate that in mutation (migration)-selection balance models the strength of apparent stabilizing selection is order V(m)/V(g). In the overdominant model and in the symmetric viability model the strength of apparent stabilizing selection is approximately 1/(2n) that of total selection on the whole phenotype. We show that a selection system that involves pairwise additive by additive epistasis in maintaining variability can lead to a lower genetic load and genetic variance in fitness (approximately 1/(2n) times) than an equivalent selection system that involves overdominance. We show that, in the epistatic model, the apparent stabilizing selection on an additive trait can be as strong as the total selection on the whole phenotype.     

Genetic variation maintained in multilocus models of additive quantitative traits under stabilizing selection.

Stabilizing selection for an intermediate optimum is generally considered to deplete genetic variation in quantitative traits. However, conflicting results from various types of models have been obtained. While classical analyses assuming a large number of independent additive loci with individually small effects indicated that no genetic variation is preserved under stabilizing selection, several analyses of two-locus models showed the contrary. We perform a complete analysis of a generalization of Wright's two-locus quadratic-optimum model and investigate numerically the ability of quadratic stabilizing selection to maintain genetic variation in additive quantitative traits controlled by up to five loci. A statistical approach is employed by choosing randomly 4000 parameter sets (allelic effects, recombination rates, and strength of selection) for a given number of loci. For each parameter set we iterate the recursion equations that describe the dynamics of gamete frequencies starting from 20 randomly chosen initial conditions until an equilibrium is reached, record the quantities of interest, and calculate their corresponding mean values. As the number of loci increases from two to five, the fraction of the genome expected to be polymorphic declines surprisingly rapidly, and the loci that are polymorphic increasingly are those with small effects on the trait. As a result, the genetic variance expected to be maintained under stabilizing selection decreases very rapidly with increased number of loci. The equilibrium structure expected under stabilizing selection on an additive trait differs markedly from that expected under selection with no constraints on genotypic fitness values. The expected genetic variance, the expected polymorphic fraction of the genome, as well as other quantities of interest, are only weakly dependent on the selection intensity and the level of recombination.