Effect of population stratification on case-control association studies. I. Elevation in false positive rates and comparison to confounding risk ratios (a simulation study)

OBJECTIVES: This is the first of two articles discussing the effect of population stratification on the type I error rate (i.e., false positive rate). This paper focuses on the confounding risk ratio (CRR). It is accepted that population stratification (PS) can produce false positive results in case-control genetic association. However, which values of population parameters lead to an increase in type I error rate is unknown. Some believe PS does not represent a serious concern, whereas others believe that PS may contribute to contradictory findings in genetic association. We used computer simulations to estimate the effect of PS on type I error rate over a wide range of disease frequencies and marker allele frequencies, and we compared the observed type I error rate to the magnitude of the confounding risk ratio. METHODS: We simulated two populations and mixed them to produce a combined population, specifying 160 different combinations of input parameters (disease prevalences and marker allele frequencies in the two populations). From the combined populations, we selected 5000 case-control datasets, each with either 50, 100, or 300 cases and controls, and determined the type I error rate. In all simulations, the marker allele and disease were independent (i.e., no association). RESULTS: The type I error rate is not substantially affected by changes in the disease prevalence per se. We found that the CRR provides a relatively poor indicator of the magnitude of the increase in type I error rate. We also derived a simple mathematical quantity, Delta, that is highly correlated with the type I error rate. In the companion article (part II, in this issue), we extend this work to multiple subpopulations and unequal sampling proportions. CONCLUSION: Based on these results, realistic combinations of disease prevalences and marker allele frequencies can substantially increase the probability of finding false evidence of marker disease associations. Furthermore, the CRR does not indicate when this will occur.     

Quantitative similarity-based association tests using population samples

Although genetic association studies using unrelated individuals may be subject to bias caused by population stratification, alternative methods that are robust to population stratification, such as family-based association designs, may be less powerful. Furthermore, it is often more feasible and less expensive to collect unrelated individuals. Recently, several statistical methods have been proposed for case-control association tests in a structured population; these methods may be robust to population stratification. In the present study, we propose a quantitative similarity-based association test (QSAT) to identify association between a candidate marker and a quantitative trait of interest, through use of unrelated individuals. For the QSAT, we first determine whether two individuals are from the same subpopulation or from different subpopulations, using genotype data at a set of independent markers. We then perform an association test between the candidate marker and the quantitative trait, through incorporation of such information. Simulation results based on either coalescent models or empirical population genetics data show that the QSAT has a correct type I error rate in the presence of population stratification and that the power of the QSAT is higher than that of family-based association designs.     

A note on exact tests of Hardy-Weinberg equilibrium

Deviations from Hardy-Weinberg equilibrium (HWE) can indicate inbreeding, population stratification, and even problems in genotyping. In samples of affected individuals, these deviations can also provide evidence for association. Tests of HWE are commonly performed using a simple chi2 goodness-of-fit test. We show that this chi2 test can have inflated type I error rates, even in relatively large samples (e.g., samples of 1,000 individuals that include approximately 100 copies of the minor allele). On the basis of previous work, we describe exact tests of HWE together with efficient computational methods for their implementation. Our methods adequately control type I error in large and small samples and are computationally efficient. They have been implemented in freely available code that will be useful for quality assessment of genotype data and for the detection of genetic association or population stratification in very large data sets.     

Simultaneously correcting for population stratification and for genotyping error in case-control association studies

In population-based case-control association studies, the regular chi (2) test is often used to investigate association between a candidate locus and disease. However, it is well known that this test may be biased in the presence of population stratification and/or genotyping error. Unlike some other biases, this bias will not go away with increasing sample size. On the contrary, the false-positive rate will be much larger when the sample size is increased. The usual family-based designs are robust against population stratification, but they are sensitive to genotype error. In this article, we propose a novel method of simultaneously correcting for the bias arising from population stratification and/or for the genotyping error in case-control studies. The appropriate corrections depend on sample odds ratios of the standard 2x3 tables of genotype by case and control from null loci. Therefore, the test is simple to apply. The corrected test is robust against misspecification of the genetic model. If the null hypothesis of no association is rejected, the corrections can be further used to estimate the effect of the genetic factor. We considered a simulation study to investigate the performance of the new method, using parameter values similar to those found in real-data examples. The results show that the corrected test approximately maintains the expected type I error rate under various simulation conditions. It also improves the power of the association test in the presence of population stratification and/or genotyping error. The discrepancy in power between the tests with correction and those without correction tends to be more extreme as the magnitude of the bias becomes larger. Therefore, the bias-correction method proposed in this article should be useful for the genetic analysis of complex traits.     

A semiparametric test to detect associations between quantitative traits and candidate genes in structured populations

MOTIVATION: Although population-based association mapping may be subject to the bias caused by population stratification, alternative methods that are robust to population stratification such as family-based linkage analysis have lower mapping resolution. Recently, various statistical methods robust to population stratification were proposed for association studies, using unrelated individuals to identify associations between candidate genes and traits of interest. The association between a candidate gene and a quantitative trait is often evaluated via a regression model with inferred population structure variables as covariates, where the residual distribution is customarily assumed to be from a symmetric and unimodal parametric family, such as a Gaussian, although this may be inappropriate for the analysis of many real-life datasets. RESULTS: In this article, we proposed a new structured association (SA) test. Our method corrects for continuous population stratification by first deriving population structure and kinship matrices through a set of random genetic markers and then modeling the relationship between trait values, genotypic scores at a candidate marker and genetic background variables through a semiparametric model, where the error distribution is modeled as a mixture of Polya trees centered around a normal family of distributions. We compared our model to the existing SA tests in terms of model fit, type I error rate, power, precision and accuracy by application to a real dataset as well as simulated datasets.     

A simple and robust TDT-type test against genotyping error with error rates varying across families

The transmission/disequilibrium test (TDT), a family based test of linkage and association, is a popular test for studies of complex inheritance, as it is nonparametric and robust against spurious conclusions induced by hidden genetic structure, such as stratification or admixture. However, the TDT may be biased by genotyping errors. Undetected genotyping errors may be contributing to an inflated type I error rate among reported TDT-derived associations. To adjust for bias, a popular approach is to assume a genotype error model for describing the pattern of errors and propose association tests using likelihood method. However, all model-based approaches tend to perform unsatisfactorily if the related genotyping error rates are not identical across all families. In this paper, we propose a TDT-type association test which is not only simple, robust against population stratification (and hence the assumption of Hardy-Weinberg equilibrium is not required), but also robust against genotyping error with error rates varying across families. Simulation studies confirm that the new test has very reasonable performance.     

Correcting for measurement error in individual ancestry estimates in structured association tests

We present theoretical explanations and show through simulation that the individual admixture proportion estimates obtained by using ancestry informative markers should be seen as an error-contaminated measurement of the underlying individual ancestry proportion. These estimates can be used in structured association tests as a control variable to limit type I error inflation or reduce loss of power due to population stratification observed in studies of admixed populations. However, the inclusion of such error-containing variables as covariates in regression models can bias parameter estimates and reduce ability to control for the confounding effect of admixture in genetic association tests. Measurement error correction methods offer a way to overcome this problem but require an a priori estimate of the measurement error variance. We show how an upper bound of this variance can be obtained, present four measurement error correction methods that are applicable to this problem, and conduct a simulation study to compare their utility in the case where the admixed population results from the intermating between two ancestral populations. Our results show that the quadratic measurement error correction (QMEC) method performs better than the other methods and maintains the type I error to its nominal level.     

Comparison of three methods to estimate genetic ancestry and control for stratification in genetic association studies among admixed populations

Population stratification may confound the results of genetic association studies among unrelated individuals from admixed populations. Several methods have been proposed to estimate the ancestral information in admixed populations and used to adjust the population stratification in genetic association tests. We evaluate the performances of three different methods: maximum likelihood estimation, ADMIXMAP and Structure through various simulated data sets and real data from Latino subjects participating in a genetic study of asthma. All three methods provide similar information on the accuracy of ancestral estimates and control type I error rate at an approximately similar rate. The most important factor in determining accuracy of the ancestry estimate and in minimizing type I error rate is the number of markers used to estimate ancestry. We demonstrate that approximately 100 ancestry informative markers (AIMs) are required to obtain estimates of ancestry that correlate with correlation coefficients more than 0.9 with the true individual ancestral proportions. In addition, after accounting for the ancestry information in association tests, the excess of type I error rate is controlled at the 5% level when 100 markers are used to estimate ancestry. However, since the effect of admixture on the type I error rate worsens with sample size, the accuracy of ancestry estimates also needs to increase to make the appropriate correction. Using data from the Latino subjects, we also apply these methods to an association study between body mass index and 44 AIMs. These simulations are meant to provide some practical guidelines for investigators conducting association studies in admixed populations.     

Comparison of population-based association study methods correcting for population stratification

Population stratification can cause spurious associations in population-based association studies. Several statistical methods have been proposed to reduce the impact of population stratification on population-based association studies. We simulated a set of stratified populations based on the real haplotype data from the HapMap ENCODE project, and compared the relative power, type I error rates, accuracy and positive prediction value of four prevailing population-based association study methods: traditional case-control tests, structured association (SA), genomic control (GC) and principal components analysis (PCA) under various population stratification levels. Additionally, we evaluated the effects of sample sizes and frequencies of disease susceptible allele on the performance of the four analytical methods in the presence of population stratification. We found that the performance of PCA was very stable under various scenarios. Our comparison results suggest that SA and PCA have comparable performance, if sufficient ancestral informative markers are used in SA analysis. GC appeared to be strongly conservative in significantly stratified populations. It may be better to apply GC in the stratified populations with low stratification level. Our study intends to provide a practical guideline for researchers to select proper study methods and make appropriate inference of the results in population-based association studies.     

A Permutation Test for Oligoset DNA Pooling Studies

Case-control association studies often suffer from population stratification bias. A previous triple combination strategy of stratum matching, genomic controlling, and multiple DNA pooling can correct the bias and save genotyping cost. However the method requires researchers to prepare a multitude of DNA pools—more than 30 case-control pooling sets in total (polyset). In this paper, the authors propose a permutation test for oligoset DNA pooling studies. Monte-Carlo simulations show that the proposed test has a type I error rate under control and a power comparable to that of individual genotyping. For a researcher on a tight budget, oligoset DNA pooling is a viable option.     

Genetic association analysis using sibship data: a multilevel model approach

Family based association study (FBAS) has the advantages of controlling for population stratification and testing for linkage and association simultaneously. We propose a retrospective multilevel model (rMLM) approach to analyze sibship data by using genotypic information as the dependent variable. Simulated data sets were generated using the simulation of linkage and association (SIMLA) program. We compared rMLM to sib transmission/disequilibrium test (S-TDT), sibling disequilibrium test (SDT), conditional logistic regression (CLR) and generalized estimation equations (GEE) on the measures of power, type I error, estimation bias and standard error. The results indicated that rMLM was a valid test of association in the presence of linkage using sibship data. The advantages of rMLM became more evident when the data contained concordant sibships. Compared to GEE, rMLM had less underestimated odds ratio (OR). Our results support the application of rMLM to detect gene-disease associations using sibship data. However, the risk of increasing type I error rate should be cautioned when there is association without linkage between the disease locus and the genotyped marker.     

Testing gene-gene interactions in the case-parents design

The case-parents design has been widely used to detect genetic associations as it can prevent spurious association that could occur in population-based designs. When examining the effect of an individual genetic locus on a disease, logistic regressions developed by conditioning on parental genotypes provide complete protection from spurious association caused by population stratification. However, when testing gene-gene interactions, it is unknown whether conditional logistic regressions are still robust. Here we evaluate the robustness and efficiency of several gene-gene interaction tests that are derived from conditional logistic regressions. We found that in the presence of SNP genotype correlation due to population stratification or linkage disequilibrium, tests with incorrectly specified main-genetic-effect models can lead to inflated type I error rates. We also found that a test with fully flexible main genetic effects always maintains correct test size and its robustness can be achieved with negligible sacrifice of its power. When testing gene-gene interactions is the focus, the test allowing fully flexible main effects is recommended to be used.     

On the Prospects of Whole-Genome Association Mapping in Saccharomyces cerevisiae

Advances in sequencing technology have enabled whole-genome sequences to be obtained from multiple individuals within species, particularly in model organisms with compact genomes. For example, 36 genome sequences of Saccharomyces cerevisiae are now publicly available, and SNP data are available for even larger collections of strains. One potential use of these resources is mapping the genetic basis of phenotypic variation through genome-wide association (GWA) studies, with the benefit that associated variants can be studied experimentally with greater ease than in outbred populations such as humans. Here, we evaluate the prospects of GWA studies in S. cerevisiae strains through extensive simulations and a GWA study of mitochondrial copy number. We demonstrate that the complex and heterogeneous patterns of population structure present in yeast populations can lead to a high type I error rate in GWA studies of quantitative traits, and that methods typically used to control for population stratification do not provide adequate control of the type I error rate. Moreover, we show that while GWA studies of quantitative traits in S. cerevisiae may be difficult depending on the particular set of strains studied, association studies to map cis-acting quantitative trait loci (QTL) and Mendelian phenotypes are more feasible. We also discuss sampling strategies that could enable GWA studies in yeast and illustrate the utility of this approach in Saccharomyces paradoxus. Thus, our results provide important practical insights into the design and interpretation of GWA studies in yeast, and other model organisms that possess complex patterns of population structure.     

Control of population stratification by correlation-selected principal components

In genome-wide association studies, population stratification is recognized as producing inflated type I error due to the inflation of test statistics. Principal component-based methods applied to genotypes provide information about population structure, and have been widely used to control for stratification. Here we explore the precise relationship between genotype principal components and inflation of association test statistics, thereby drawing a connection between principal component-based stratification control and the alternative approach of genomic control. Our results provide an inherent justification for the use of principal components, but call into question the popular practice of selecting principal components based on significance of eigenvalues alone. We propose a new approach, called EigenCorr, which selects principal components based on both their eigenvalues and their correlation with the (disease) phenotype. Our approach tends to select fewer principal components for stratification control than does testing of eigenvalues alone, providing substantial computational savings and improvements in power. Analyses of simulated and real data demonstrate the usefulness of the proposed approach.     

The bias introduced by population stratification in IBD based linkage analysis

The lack of replication of model-free linkage analyses performed on complex diseases raises questions about the robustness of these methods to various biases. The confounding effect of population stratification on a genetic association study has long been recognized in the genetic epidemiology community. Because the estimation of the number of alleles shared identical by descent (IBD) does not depend on the marker allele frequency when founders of families are observed, model-free linkage analysis is usually thought to be robust to population stratification. However, for common complex diseases, the genotypes of founders are often unobserved and therefore population stratification has the potential to impair model-free linkage analysis. Here, we demonstrate that, when some or all of the founder genotypes are missing, population stratification can introduce deleterious effects on various model-free linkage methods or designs. For an affected sib pair design, it can cause excess false-positive discoveries even when the trait distribution is homogeneous among subpopulations. After incorporating a control group of discordant sib pairs or for a quantitative trait, two circumstances must be met for population stratification to be a confounder: the distributions for both the marker and the trait must be heterogeneous among subpopulations. When this occurs, the bias can result in either a liberal, and hence invalid, test or a conservative test. Bias can be eliminated or alleviated by inclusion of founders' or other family members' genotype data. When this is not possible, new methods need to be developed to be robust to population stratification.     

Admixture-matched case-control study: a practical approach for genetic association studies in admixed populations

Case-control genetic association studies in admixed populations are known to be susceptible to genetic confounding due to population stratification. The transmission/disequilibrium test (TDT) approach can avoid this problem. However, the TDT is expensive and impractical for late-onset diseases. Case-control study designs, in which, cases and controls are matched by admixture, can be an appealing and a suitable alternative for genetic association studies in admixed populations. In this study, we applied this matching strategy when recruiting our African American participants in the Study of African American, Asthma, Genes and Environments. Group admixture in this cohort consists of 83% African ancestry and 17% European ancestry, which was consistent with reports from other studies. By carrying out several complementary analyses, our results show that there is a substructure in the cohort, but that the admixture distributions are almost identical in cases and controls, and also in cases only. We performed association tests for asthma-related traits with ancestry, and only found that FEV(1), a measure for baseline pulmonary function, was associated with ancestry after adjusting for socio-economic and environmental risk factors (P=0.01). We did not observe an excess of type I error rate in our association tests for ancestry informative markers and asthma-related phenotypes when ancestry was not adjusted in the analyses. Furthermore, using the association tests between genetic variants in a known asthma candidate gene, beta(2) adrenergic receptor (beta(2)AR) and DeltaFEF(25-75), an asthma-related phenotype, as an example, we demonstrated population stratification was not a confounder in our genetic association. Our present work demonstrates that admixture-matched case-control strategies can efficiently control population stratification confounding in admixed populations.     

Adjustment for local ancestry in genetic association analysis of admixed populations

MOTIVATION: Admixed populations offer a unique opportunity for mapping diseases that have large disease allele frequency differences between ancestral populations. However, association analysis in such populations is challenging because population stratification may lead to association with loci unlinked to the disease locus. Methods and results: We show that local ancestry at a test single nucleotide polymorphism (SNP) may confound with the association signal and ignoring it can lead to spurious association. We demonstrate theoretically that adjustment for local ancestry at the test SNP is sufficient to remove the spurious association regardless of the mechanism of population stratification, whether due to local or global ancestry differences among study subjects; however, global ancestry adjustment procedures may not be effective. We further develop two novel association tests that adjust for local ancestry. Our first test is based on a conditional likelihood framework which models the distribution of the test SNP given disease status and flanking marker genotypes. A key advantage of this test lies in its ability to incorporate different directions of association in the ancestral populations. Our second test, which is computationally simpler, is based on logistic regression, with adjustment for local ancestry proportion. We conducted extensive simulations and found that the Type I error rates of our tests are under control; however, the global adjustment procedures yielded inflated Type I error rates when stratification is due to local ancestry difference.     

Using Network Methodology to Infer Population Substructure

One of the main caveats of association studies is the possible affection by bias due to population stratification. Existing methods rely on model-based approaches like structure and ADMIXTURE or on principal component analysis like EIGENSTRAT. Here we provide a novel visualization technique and describe the problem of population substructure from a graph-theoretical point of view. We group the sequenced individuals into triads, which depict the relational structure, on the basis of a predefined pairwise similarity measure. We then merge the triads into a network and apply community detection algorithms in order to identify homogeneous subgroups or communities, which can further be incorporated as covariates into logistic regression. We apply our method to populations from different continents in the 1000 Genomes Project and evaluate the type 1 error based on the empirical p-values. The application to 1000 Genomes data suggests that the network approach provides a very fine resolution of the underlying ancestral population structure. Besides we show in simulations, that in the presence of discrete population structures, our developed approach maintains the type 1 error more precisely than existing approaches.     

A Monte Carlo Permutation Test for Random Mating Using Genome Sequences

Testing for random mating of a population is important in population genetics, because deviations from randomness of mating may indicate inbreeding, population stratification, natural selection, or sampling bias. However, current methods use only observed numbers of genotypes and alleles, and do not take advantage of the fact that the advent of sequencing technology provides an opportunity to investigate this topic in unprecedented detail. To address this opportunity, a novel statistical test for random mating is required in population genomics studies for which large sequencing datasets are generally available. Here, we propose a Monte-Carlo-based-permutation test (MCP) as an approach to detect random mating. Computer simulations used to evaluate the performance of the permutation test indicate that its type I error is well controlled and that its statistical power is greater than that of the commonly used chi-square test (CHI). Our simulation study shows the power of our test is greater for datasets characterized by lower levels of migration between subpopulations. In addition, test power increases with increasing recombination rate, sample size, and divergence time of subpopulations. For populations exhibiting limited migration and having average levels of population divergence, the statistical power approaches 1 for sequences longer than 1Mbp and for samples of 400 individuals or more. Taken together, our results suggest that our permutation test is a valuable tool to detect random mating of populations, especially in population genomics studies.     

Comparison of tests for association and linkage in incomplete families

To analyze incomplete families, the following statistical tests can be used: LRAT-a simple likelihood-based association test, TRANSMIT, SIBASSOC/STDT, and RCTDT. We compared these four tests, for the diallelic case, on simulated data sets. The comparisons focused on the power to detect linkage and association when different familial structures, resistance to population stratification, resistance to misclassification of the disease status of the healthy sib, and the effect of nonpaternity were considered. The simulations lead to the following conclusions. The type I errors of TRANSMIT, SIBASSOC/STDT, and RCTDT were not affected by population stratification. LRAT showed bias under strong population stratification. High nonpaternity rates can lead to inflated type I errors, highlighting the importance of identification of half sibs. Under different homogeneous models, the power of TRANSMIT was very similar to that of LRAT, and, similarly, no difference in power was observed between SIBASSOC/STDT and RCTDT. Under various recessive and additive models, TRANSMIT was slightly more powerful than SIBASSOC/STDT when monoparental families with one affected and one unaffected sib were analyzed. Under various dominant models, SIBASSOC/STDT was slightly more powerful than TRANSMIT. Misclassification of the disease status of healthy sibs, as well as the discarding of incomplete families, resulted in a consistent loss of power.