Haplotype Inference for Population Data with Genotyping Errors

Inference of haplotypes is important in genetic epidemiology studies. However, all large genotype data sets have errors due to the use of inexpensive genotyping machines that are fallible and shortcomings in genotyping scoring softwares, which can have an enormous impact on haplotype inference. In this article, we propose two novel strategies to reduce the impact induced by genotyping errors in haplotype inference. The first method makes use of double sampling. For each individual, the “GenoSpectrum” that consists of all possible genotypes and their corresponding likelihoods are computed. The second method is a genotype clustering algorithm based on multi‐genotyping data, which also assigns a “GenoSpectrum” for each individual. We then describe two hybrid EM algorithms (called DS‐EM and MG‐EM) that perform haplotype inference based on “GenoSpectrum” of each individual obtained by double sampling and multi‐genotyping data. Both simulated data sets and a quasi real‐data set demonstrate that our proposed methods perform well in different situations and outperform the conventional EM algorithm and the HMM algorithm proposed by Sun, Greenwood, and Neal (2007, Genetic Epidemiology 31 , 937–948) when the genotype data sets have errors.     

A new method for haplotype inference including full-sib information

Recent literature has suggested that haplotype inference through close relatives, especially from nuclear families, can be an alternative strategy in determining linkage phase and estimating haplotype frequencies. In the case of no possibility to obtain genotypes for parents, and only full-sib information being used, a new approach is suggested to infer phase and to reconstruct haplotypes. We present a maximum-likelihood method via an expectation-maximization algorithm, called FSHAP, using only full-sib information when parent information is not available. FSHAP can deal with families with an arbitrary number of children, and missing parents or missing genotypes can be handled as well. In a simulation study we compare FSHAP with another existing expectation-maximization (EM)-based approach (FAMHAP), the conditioning approach implemented in FBAT and GENEHUNTER, which is only pedigree based and assumes linkage equilibrium. In most situations, FSHAP has the smallest discrepancy of haplotype frequency estimation and the lowest error rate in haplotype reconstruction, only in some cases FAMHAP yields comparable results. GENEHUNTER produces the largest discrepancy, and FBAT produces the highest error rate in offspring in most situations. Among the methods compared, FSHAP has the highest accuracy in reconstructing the diplotypes of the unavailable parents. Potential limitations of the method, e.g., in analyzing very large haplotypes, are indicated and possible solutions are discussed.     

Bayesian haplotype inference via the Dirichlet process.

The problem of inferring haplotypes from genotypes of single nucleotide polymorphisms (SNPs) is essential for the understanding of genetic variation within and among populations, with important applications to the genetic analysis of disease propensities and other complex traits. The problem can be formulated as a mixture model, where the mixture components correspond to the pool of haplotypes in the population. The size of this pool is unknown; indeed, knowing the size of the pool would correspond to knowing something significant about the genome and its history. Thus methods for fitting the genotype mixture must crucially address the problem of estimating a mixture with an unknown number of mixture components. In this paper we present a Bayesian approach to this problem based on a nonparametric prior known as the Dirichlet process. The model also incorporates a likelihood that captures statistical errors in the haplotype/genotype relationship trading off these errors against the size of the pool of haplotypes. We describe an algorithm based on Markov chain Monte Carlo for posterior inference in our model. The overall result is a flexible Bayesian method, referred to as DP-Haplotyper, that is reminiscent of parsimony methods in its preference for small haplotype pools. We further generalize the model to treat pedigree relationships (e.g., trios) between the population's genotypes. We apply DP-Haplotyper to the analysis of both simulated and real genotype data, and compare to extant methods.     

Accurate haplotype inference for multiple linked single-nucleotide polymorphisms using sibship data

Sibships are commonly used in genetic dissection of complex diseases, particularly for late-onset diseases. Haplotype-based association studies have been advocated as powerful tools for fine mapping and positional cloning of complex disease genes. Existing methods for haplotype inference using data from relatives were originally developed for pedigree data. In this study, we proposed a new statistical method for haplotype inference for multiple tightly linked single-nucleotide polymorphisms (SNPs), which is tailored for extensively accumulated sibship data. This new method was implemented via an expectation-maximization (EM) algorithm without the usual assumption of linkage equilibrium among markers. Our EM algorithm does not incur extra computational burden for haplotype inference using sibship data when compared with using unrelated parental data. Furthermore, its computational efficiency is not affected by increasing sibship size. We examined the robustness and statistical performance of our new method in simulated data created from an empirical haplotype data set of human growth hormone gene 1. The utility of our method was illustrated with an application to the analyses of haplotypes of three candidate genes for osteoporosis.     

2SNP: scalable phasing based on 2-SNP haplotypes

2SNP software package implements a new very fast scalable algorithm for haplotype inference based on genotype statistics collected only for pairs of SNPs. This software can be used for comparatively accurate phasing of large number of long genome sequences, e.g. obtained from DNA arrays. As an input 2SNP takes genotype matrix and outputs the corresponding haplotype matrix. On datasets across 79 regions from HapMap 2SNP is several orders of magnitude faster than GERBIL and PHASE while matching them in quality measured by the number of correctly phased genotypes, single-site and switching errors. For example, 2SNP requires 41 s on Pentium 4 2 Ghz processor to phase 30 genotypes with 1381 SNPs (ENm010.7p15:2 data from HapMap) versus GERBIL and PHASE requiring more than a week and admitting no less errors than 2SNP.     

Inference on haplotype effects in case-control studies using unphased genotype data

A variety of statistical methods exist for detecting haplotype-disease association through use of genetic data from a case-control study. Since such data often consist of unphased genotypes (resulting in haplotype ambiguity), such statistical methods typically apply the expectation-maximization (EM) algorithm for inference. However, the majority of these methods fail to perform inference on the effect of particular haplotypes or haplotype features on disease risk. Since such inference is valuable, we develop a retrospective likelihood for estimating and testing the effects of specific features of single-nucleotide polymorphism (SNP)-based haplotypes on disease risk using unphased genotype data from a case-control study. Our proposed method has a flexible structure that allows, among other choices, modeling of multiplicative, dominant, and recessive effects of specific haplotype features on disease risk. In addition, our method relaxes the requirement of Hardy-Weinberg equilibrium of haplotype frequencies in case subjects, which is typically required of EM-based haplotype methods. Also, our method easily accommodates missing SNP information. Finally, our method allows for asymptotic, permutation-based, or bootstrap inference. We apply our method to case-control SNP genotype data from the Finland-United States Investigation of Non-Insulin-Dependent Diabetes Mellitus (FUSION) Genetics study and identify two haplotypes that appear to be significantly associated with type 2 diabetes. Using the FUSION data, we assess the accuracy of asymptotic P values by comparing them with P values obtained from a permutation procedure. We also assess the accuracy of asymptotic confidence intervals for relative-risk parameters for haplotype effects, by a simulation study based on the FUSION data.     

Haplotype reconstruction for diploid populations

The inference of haplotype pairs directly from unphased genotype data is a key step in the analysis of genetic variation in relation to disease and pharmacogenetically relevant traits. Most popular methods such as Phase and PL do require either the coalescence assumption or the assumption of linkage between the single-nucleotide polymorphisms (SNPs). We have now developed novel approaches that are independent of these assumptions. First, we introduce a new optimization criterion in combination with a block-wise evolutionary Monte Carlo algorithm. Based on this criterion, the 'haplotype likelihood', we develop two kinds of estimators, the maximum haplotype-likelihood (MHL) estimator and its empirical Bayesian (EB) version. Using both real and simulated data sets, we demonstrate that our proposed estimators allow substantial improvements over both the expectation-maximization (EM) algorithm and Clark's procedure in terms of capacity/scalability and error rate. Thus, hundreds and more ambiguous loci and potentially very large sample sizes can be processed. Moreover, applying our proposed EB estimator can result in significant reductions of error rate in the case of unlinked or only weakly linked SNPs.     

A statistical framework for haplotype block inference

The existence of haplotype blocks transmitted from parents to offspring has been suggested recently. This has created an interest in the inference of the block structure and length. The motivation is that haplotype blocks that are characterized well will make it relatively easier to quickly map all the genes carrying human diseases. To study the inference of haplotype block systematically, we propose a statistical framework. In this framework, the optimal haplotype block partitioning is formulated as the problem of statistical model selection; missing data can be handled in a standard statistical way; population strata can be implemented; block structure inference/hypothesis testing can be performed; prior knowledge, if present, can be incorporated to perform a Bayesian inference. The algorithm is linear in the number of loci, instead of NP-hard for many such algorithms. We illustrate the applications of our method to both simulated and real data sets.     

Maximum likelihood estimation of linkage disequilibrium in half-sib families

Maximum likelihood methods for the estimation of linkage disequilibrium between biallelic DNA-markers in half-sib families (half-sib method) are developed for single and multifamily situations. Monte Carlo computer simulations were carried out for a variety of scenarios regarding sire genotypes, linkage disequilibrium, recombination fraction, family size, and number of families. A double heterozygote sire was simulated with recombination fraction of 0.00, linkage disequilibrium among dams of δ=0.10, and alleles at both markers segregating at intermediate frequencies for a family size of 500. The average estimates of δ were 0.17, 0.25, and 0.10 for Excoffier and Slatkin (1995), maternal informative haplotypes, and the half-sib method, respectively. A multifamily EM algorithm was tested at intermediate frequencies by computer simulation. The range of the absolute difference between estimated and simulated δ was between 0.000 and 0.008. A cattle half-sib family was genotyped with the Illumina 50K BeadChip. There were 314,730 SNP pairs for which the sire was a homo-heterozygote with average estimates of r2 of 0.115, 0.067, and 0.111 for half-sib, Excoffier and Slatkin (1995), and maternal informative haplotypes methods, respectively. There were 208,872 SNP pairs for which the sire was double heterozygote with average estimates of r2 across the genome of 0.100, 0.267, and 0.925 for half-sib, Excoffier and Slatkin (1995), and maternal informative haplotypes methods, respectively. Genome analyses for all possible sire genotypes with 829,042 tests showed that ignoring half-sib family structure leads to upward biased estimates of linkage disequilibrium. Published inferences on population structure and evolution of cattle should be revisited after accommodating existing half-sib family structure in the estimation of linkage disequilibrium.     

The impact of missing and erroneous genotypes on tagging SNP selection and power of subsequent association tests

OBJECTIVE: Single nucleotide polymorphisms (SNPs) serve as effective markers for localizing disease susceptibility genes, but current genotyping technologies are inadequate for genotyping all available SNP markers in a typical linkage/association study. Much attention has recently been paid to methods for selecting the minimal informative subset of SNPs in identifying haplotypes, but there has been little investigation of the effect of missing or erroneous genotypes on the performance of these SNP selection algorithms and subsequent association tests using the selected tagging SNPs. The purpose of this study is to explore the effect of missing genotype or genotyping error on tagging SNP selection and subsequent single marker and haplotype association tests using the selected tagging SNPs. METHODS: Through two sets of simulations, we evaluated the performance of three tagging SNP selection programs in the presence of missing or erroneous genotypes: Clayton's diversity based program htstep, Carlson's linkage disequilibrium (LD) based program ldSelect, and Stram's coefficient of determination based program tagsnp.exe. RESULTS: When randomly selected known loci were relabeled as 'missing', we found that the average number of tagging SNPs selected by all three algorithms changed very little and the power of subsequent single marker and haplotype association tests using the selected tagging SNPs remained close to the power of these tests in the absence of missing genotype. When random genotyping errors were introduced, we found that the average number of tagging SNPs selected by all three algorithms increased. In data sets simulated according to the haplotype frequecies in the CYP19 region, Stram's program had larger increase than Carlson's and Clayton's programs. In data sets simulated under the coalescent model, Carlson's program had the largest increase and Clayton's program had the smallest increase. In both sets of simulations, with the presence of genotyping errors, the power of the haplotype tests from all three programs decreased quickly, but there was not much reduction in power of the single marker tests. CONCLUSIONS: Missing genotypes do not seem to have much impact on tagging SNP selection and subsequent single marker and haplotype association tests. In contrast, genotyping errors could have severe impact on tagging SNP selection and haplotype tests, but not on single marker tests.     

Haplotype inference for present-absent genotype data using previously identified haplotypes and haplotype patterns

MOTIVATION: Killer immunoglobulin-like receptor (KIR) genes vary considerably in their presence or absence on a specific regional haplotype. Because presence or absence of these genes is largely detected using locus-specific genotyping technology, the distinction between homozygosity and hemizygosity is often ambiguous. The performance of methods for haplotype inference (e.g. PL-EM, PHASE) for KIR genes may be compromised due to the large portion of ambiguous data. At the same time, many haplotypes or partial haplotype patterns have been previously identified and can be incorporated to facilitate haplotype inference for unphased genotype data. To accommodate the increased ambiguity of present-absent genotyping of KIR genes, we developed a hybrid approach combining a greedy algorithm with the Expectation-Maximization (EM) method for haplotype inference based on previously identified haplotypes and haplotype patterns. RESULTS: We implemented this algorithm in a software package named HAPLO-IHP (Haplotype inference using identified haplotype patterns) and compared its performance with that of HAPLORE and PHASE on simulated KIR genotypes. We compared five measures in order to evaluate the reliability of haplotype assignments and the accuracy in estimating haplotype frequency. Our method outperformed the two existing techniques by all five measures when either 60% or 25% of previously identified haplotypes were incorporated into the analyses. AVAILABILITY: The HAPLO-IHP is available at http://www.soph.uab.edu/Statgenetics/People/KZhang/HAPLO-IHP/index.html. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.     

Inference of missing SNPs and information quantity measurements for haplotype blocks

MOTIVATION: Missing data in genotyping single nucleotide polymorphism (SNP) spots are common. High-throughput genotyping methods usually have a high rate of missing data. For example, the published human chromosome 21 data by Patil et al. contains about 20% missing SNPs. Inferring missing SNPs using the haplotype block structure is promising but difficult because the haplotype block boundaries are not well defined. Here we propose a global algorithm to overcome this difficulty. RESULTS: First, we propose to use entropy as a measure of haplotype diversity. We show that the entropy measure combined with a dynamic programming algorithm produces better haplotype block partitions than other measures. Second, based on the entropy measure, we propose a two-step iterative partition-inference algorithm for the inference of missing SNPs. At the first step, we apply the dynamic programming algorithm to partition haplotypes into blocks. At the second step, we use an iterative process similar to the expectation-maximization algorithm to infer missing SNPs in each haplotype block so as to minimize the block entropy. The algorithm iterates these two steps until the total block entropy is minimized. We test our algorithm in several experimental data sets. The results show that the global approach significantly improves the accuracy of the inference. AVAILABILITY: Upon request.     

A fast algorithm for Bayesian multi-locus model in genome-wide association studies

Genome-wide association studies (GWAS) have identified a large amount of single-nucleotide polymorphisms (SNPs) associated with complex traits. A recently developed linear mixed model for estimating heritability by simultaneously fitting all SNPs suggests that common variants can explain a substantial fraction of heritability, which hints at the low power of single variant analysis typically used in GWAS. Consequently, many multi-locus shrinkage models have been proposed under a Bayesian framework. However, most use Markov Chain Monte Carlo (MCMC) algorithm, which are time-consuming and challenging to apply to GWAS data. Here, we propose a fast algorithm of Bayesian adaptive lasso using variational inference (BAL-VI). Extensive simulations and real data analysis indicate that our model outperforms the well-known Bayesian lasso and Bayesian adaptive lasso models in accuracy and speed. BAL-VI can complete a simultaneous analysis of a lung cancer GWAS data with ~3400 subjects and ~570,000 SNPs in about half a day.     

HAPLORE: a program for haplotype reconstruction in general pedigrees without recombination

MOTIVATION: Haplotype reconstruction is an essential step in genetic linkage and association studies. Although many methods have been developed to estimate haplotype frequencies and reconstruct haplotypes for a sample of unrelated individuals, haplotype reconstruction in large pedigrees with a large number of genetic markers remains a challenging problem. METHODS: We have developed an efficient computer program, HAPLORE (HAPLOtype REconstruction), to identify all haplotype sets that are compatible with the observed genotypes in a pedigree for tightly linked genetic markers. HAPLORE consists of three steps that can serve different needs in applications. In the first step, a set of logic rules is used to reduce the number of compatible haplotypes of each individual in the pedigree as much as possible. After this step, the haplotypes of all individuals in the pedigree can be completely or partially determined. These logic rules are applicable to completely linked markers and they can be used to impute missing data and check genotyping errors. In the second step, a haplotype-elimination algorithm similar to the genotype-elimination algorithms used in linkage analysis is applied to delete incompatible haplotypes derived from the first step. All superfluous haplotypes of the pedigree members will be excluded after this step. In the third step, the expectation-maximization (EM) algorithm combined with the partition and ligation technique is used to estimate haplotype frequencies based on the inferred haplotype configurations through the first two steps. Only compatible haplotype configurations with haplotypes having frequencies greater than a threshold are retained. RESULTS: We test the effectiveness and the efficiency of HAPLORE using both simulated and real datasets. Our results show that, the rule-based algorithm is very efficient for completely genotyped pedigree. In this case, almost all of the families have one unique haplotype configuration. In the presence of missing data, the number of compatible haplotypes can be substantially reduced by HAPLORE, and the program will provide all possible haplotype configurations of a pedigree under different circumstances, if such multiple configurations exist. These inferred haplotype configurations, as well as the haplotype frequencies estimated by the EM algorithm, can be used in genetic linkage and association studies. AVAILABILITY: The program can be downloaded from http://bioinformatics.med.yale.edu.     

Effectiveness of computational methods in haplotype prediction

Haplotype analysis has been used for narrowing down the location of disease-susceptibility genes and for investigating many population processes. Computational algorithms have been developed to estimate haplotype frequencies and to predict haplotype phases from genotype data for unrelated individuals. However, the accuracy of such computational methods needs to be evaluated before their applications can be advocated. We have experimentally determined the haplotypes at two loci, the N-acetyltransferase 2 gene ( NAT2, 850 bp, n=81) and a 140-kb region on chromosome X ( n=77), each consisting of five single nucleotide polymorphisms (SNPs). We empirically evaluated and compared the accuracy of the subtraction method, the expectation-maximization (EM) method, and the PHASE method in haplotype frequency estimation and in haplotype phase prediction. Where there was near complete linkage disequilibrium (LD) between SNPs (the NAT2 gene), all three methods provided effective and accurate estimates for haplotype frequencies and individual haplotype phases. For a genomic region in which marked LD was not maintained (the chromosome X locus), the computational methods were adequate in estimating overall haplotype frequencies. However, none of the methods was accurate in predicting individual haplotype phases. The EM and the PHASE methods provided better estimates for overall haplotype frequencies than the subtraction method for both genomic regions.     

A comparison of phasing algorithms for trios and unrelated individuals

Knowledge of haplotype phase is valuable for many analysis methods in the study of disease, population, and evolutionary genetics. Considerable research effort has been devoted to the development of statistical and computational methods that infer haplotype phase from genotype data. Although a substantial number of such methods have been developed, they have focused principally on inference from unrelated individuals, and comparisons between methods have been rather limited. Here, we describe the extension of five leading algorithms for phase inference for handling father-mother-child trios. We performed a comprehensive assessment of the methods applied to both trios and to unrelated individuals, with a focus on genomic-scale problems, using both simulated data and data from the HapMap project. The most accurate algorithm was PHASE (v2.1). For this method, the percentages of genotypes whose phase was incorrectly inferred were 0.12%, 0.05%, and 0.16% for trios from simulated data, HapMap Centre d'Etude du Polymorphisme Humain (CEPH) trios, and HapMap Yoruban trios, respectively, and 5.2% and 5.9% for unrelated individuals in simulated data and the HapMap CEPH data, respectively. The other methods considered in this work had comparable but slightly worse error rates. The error rates for trios are similar to the levels of genotyping error and missing data expected. We thus conclude that all the methods considered will provide highly accurate estimates of haplotypes when applied to trio data sets. Running times differ substantially between methods. Although it is one of the slowest methods, PHASE (v2.1) was used to infer haplotypes for the 1 million-SNP HapMap data set. Finally, we evaluated methods of estimating the value of r(2) between a pair of SNPs and concluded that all methods estimated r(2) well when the estimated value was >or=0.8.     

Estimating Haplotype Effects for Survival Data

Summary Genetic association studies often investigate the effect of haplotypes on an outcome of interest. Haplotypes are not observed directly, and this complicates the inclusion of such effects in survival models. We describe a new estimating equations approach for Cox's regression model to assess haplotype effects for survival data. These estimating equations are simple to implement and avoid the use of the EM algorithm, which may be slow in the context of the semiparametric Cox model with incomplete covariate information. These estimating equations also lead to easily computable, direct estimators of standard errors, and thus overcome some of the difficulty in obtaining variance estimators based on the EM algorithm in this setting. We also develop an easily implemented goodness‐of‐fit procedure for Cox's regression model including haplotype effects. Finally, we apply the procedures presented in this article to investigate possible haplotype effects of the PAF‐receptor on cardiovascular events in patients with coronary artery disease, and compare our results to those based on the EM algorithm.     

Accuracy of haplotype reconstruction from haplotype-tagging single-nucleotide polymorphisms

Many investigators are now using haplotype-tagging single-nucleotide polymorphism (htSNPs) as a way of screening regions of the genome for association with disease. A common approach is to genotype htSNPs in a study population and to use this information to draw inferences about each individual's haplotypic makeup, including SNPs that were not directly genotyped. To test the validity of this approach, we simulated the exercise of typing htSNPs in a large sample of individuals and compared the true and inferred haplotypes. The accuracy of haplotype inference varied, depending on the method of selecting htSNPs, the linkage-disequilibrium structure of the region, and the amount of missing data. At the stage of selection of htSNPs, haplotype-block-based methods required a larger number of htSNPs than did unstructured methods but gave lower levels of error in haplotype inference, particularly when there was a significant amount of missing data. We present a Web-based utility that allows investigators to compare the likely error rates of different sets of htSNPs and to arrive at an economical set of htSNPs that provides acceptable levels of accuracy in haplotype inference.     

Rapid and accurate haplotype phasing and missing-data inference for whole-genome association studies by use of localized haplotype clustering

Whole-genome association studies present many new statistical and computational challenges due to the large quantity of data obtained. One of these challenges is haplotype inference; methods for haplotype inference designed for small data sets from candidate-gene studies do not scale well to the large number of individuals genotyped in whole-genome association studies. We present a new method and software for inference of haplotype phase and missing data that can accurately phase data from whole-genome association studies, and we present the first comparison of haplotype-inference methods for real and simulated data sets with thousands of genotyped individuals. We find that our method outperforms existing methods in terms of both speed and accuracy for large data sets with thousands of individuals and densely spaced genetic markers, and we use our method to phase a real data set of 3,002 individuals genotyped for 490,032 markers in 3.1 days of computing time, with 99% of masked alleles imputed correctly. Our method is implemented in the Beagle software package, which is freely available.     

Comparison of the accuracy of methods of computational haplotype inference using a large empirical dataset

Background  

Analyses of genetic data at the level of haplotypes provide increased accuracy and power to infer genotype-phenotype correlations and evolutionary history of a locus. However, empirical determination of haplotypes is expensive and laborious. Therefore, several methods of inferring haplotypes from unphased genotypic data have been proposed, but it is unclear how accurate each of the methods is or which methods are superior. The accuracy of some of the leading methods of computational haplotype inference (PL-EM, Phase, SNPHAP, Haplotyper) are compared using a large set of 308 empirically determined haplotypes based on 15 SNPs, among which 36 haplotypes were observed to occur. This study presents several advantages over many previous comparisons of haplotype inference methods: a large number of subjects are included, the number of known haplotypes is much smaller than the number of chromosomes surveyed, a range in values of linkage disequilibrium, presence of rare SNP alleles, and considerable dispersion in the frequencies of haplotypes.