Evaluating and improving power in whole-genome association studies using fixed marker sets

Emerging technologies make it possible for the first time to genotype hundreds of thousands of SNPs simultaneously, enabling whole-genome association studies. Using empirical genotype data from the International HapMap Project, we evaluate the extent to which the sets of SNPs contained on three whole-genome genotyping arrays capture common SNPs across the genome, and we find that the majority of common SNPs are well captured by these products either directly or through linkage disequilibrium. We explore analytical strategies that use HapMap data to improve power of association studies conducted with these fixed sets of markers and show that limited inclusion of specific haplotype tests in association analysis can increase the fraction of common variants captured by 25-100%. Finally, we introduce a Bayesian approach to association analysis by weighting the likelihood of each statistical test to reflect the number of putative causal alleles to which it is correlated.     

Prospects for whole-genome linkage disequilibrium mapping of common disease genes.

Recently, attention has focused on the use of whole-genome linkage disequilibrium (LD) studies to map common disease genes. Such studies would employ a dense map of single nucleotide polymorphisms (SNPs) to detect association between a marker and disease. Construction of SNP maps is currently underway. An essential issue yet to be settled is the required marker density of such maps. Here, I use population simulations to estimate the extent of LD surrounding common gene variants in the general human population as well as in isolated populations. Two main conclusions emerge from these investigations. First, a useful level of LD is unlikely to extend beyond an average distance of roughly 3 kb in the general population, which implies that approximately 500,000 SNPs will be required for whole-genome studies. Second, the extent of LD is similar in isolated populations unless the founding bottleneck is very narrow or the frequency of the variant is low (<5%).     

Genomic screening and replication using the same data set in family-based association testing

The Human Genome Project and its spin-offs are making it increasingly feasible to determine the genetic basis of complex traits using genome-wide association studies. The statistical challenge of analyzing such studies stems from the severe multiple-comparison problem resulting from the analysis of thousands of SNPs. Our methodology for genome-wide family-based association studies, using single SNPs or haplotypes, can identify associations that achieve genome-wide significance. In relation to developing guidelines for our screening tools, we determined lower bounds for the estimated power to detect the gene underlying the disease-susceptibility locus, which hold regardless of the linkage disequilibrium structure present in the data. We also assessed the power of our approach in the presence of multiple disease-susceptibility loci. Our screening tools accommodate genomic control and use the concept of haplotype-tagging SNPs. Our methods use the entire sample and do not require separate screening and validation samples to establish genome-wide significance, as population-based designs do.     

Transferability of tag SNPs in genetic association studies in multiple populations

A general question for linkage disequilibrium-based association studies is how power to detect an association is compromised when tag SNPs are chosen from data in one population sample and then deployed in another sample. Specifically, it is important to know how well tags picked from the HapMap DNA samples capture the variation in other samples. To address this, we collected dense data uniformly across the four HapMap population samples and eleven other population samples. We picked tag SNPs using genotype data we collected in the HapMap samples and then evaluated the effective coverage of these tags in comparison to the entire set of common variants observed in the other samples. We simulated case-control association studies in the non-HapMap samples under a disease model of modest risk, and we observed little loss in power. These results demonstrate that the HapMap DNA samples can be used to select tags for genome-wide association studies in many samples around the world.     

Estimating coverage and power for genetic association studies using near-complete variation data

Although studies suggest that SNPs derived from HapMap provide promising coverage and power for association studies, the lack of alternative variation datasets limits independent analysis. Using near-complete variation data for 76 genes resequenced in HapMap samples, we find that coverage of common variation by commercial genotyping arrays is substantially lower compared to the HapMap-based estimates. We quantify the power offered by these arrays for a range of disease models.     

Genome-wide mapping with biallelic markers in Arabidopsis thaliana.

Single-nucleotide polymorphisms, as well as small insertions and deletions (here referred to collectively as simple nucleotide polymorphisms, or SNPs), comprise the largest set of sequence variants in most organisms. Positional cloning based on SNPs may accelerate the identification of human disease traits and a range of biologically informative mutations. The recent application of high-density oligonucleotide arrays to allele identification has made it feasible to genotype thousands of biallelic SNPs in a single experiment. It has yet to be established, however, whether SNP detection using oligonucleotide arrays can be used to accelerate the mapping of traits in diploid genomes. The cruciferous weed Arabidopsis thaliana is an attractive model system for the construction and use of biallelic SNP maps. Although important biological processes ranging from fertilization and cell fate determination to disease resistance have been modelled in A. thaliana, identifying mutations in this organism has been impeded by the lack of a high-density genetic map consisting of easily genotyped DNA markers. We report here the construction of a biallelic genetic map in A. thaliana with a resolution of 3.5 cM and its use in mapping Eds16, a gene involved in the defence response to the fungal pathogen Erysiphe orontii. Mapping of this trait involved the high-throughput generation of meiotic maps of F2 individuals using high-density oligonucleotide probe array-based genotyping. We developed a software package called InterMap and used it to automatically delimit Eds16 to a 7-cM interval on chromosome 1. These results are the first demonstration of biallelic mapping in diploid genomes and establish means for generalizing SNP-based maps to virtually any genetic organism.     

Bayesian inference of epistatic interactions in case-control studies

Epistatic interactions among multiple genetic variants in the human genome may be important in determining individual susceptibility to common diseases. Although some existing computational methods for identifying genetic interactions have been effective for small-scale studies, we here propose a method, denoted 'bayesian epistasis association mapping' (BEAM), for genome-wide case-control studies. BEAM treats the disease-associated markers and their interactions via a bayesian partitioning model and computes, via Markov chain Monte Carlo, the posterior probability that each marker set is associated with the disease. Testing this on an age-related macular degeneration genome-wide association data set, we demonstrate that the method is significantly more powerful than existing approaches and that genome-wide case-control epistasis mapping with many thousands of markers is both computationally and statistically feasible.     

Bayesian inference analyses of the polygenic architecture of rheumatoid arthritis

The genetic architectures of common, complex diseases are largely uncharacterized. We modeled the genetic architecture underlying genome-wide association study (GWAS) data for rheumatoid arthritis and developed a new method using polygenic risk-score analyses to infer the total liability-scale variance explained by associated GWAS SNPs. Using this method, we estimated that, together, thousands of SNPs from rheumatoid arthritis GWAS explain an additional 20% of disease risk (excluding known associated loci). We further tested this method on datasets for three additional diseases and obtained comparable estimates for celiac disease (43% excluding the major histocompatibility complex), myocardial infarction and coronary artery disease (48%) and type 2 diabetes (49%). Our results are consistent with simulated genetic models in which hundreds of associated loci harbor common causal variants and a smaller number of loci harbor multiple rare causal variants. These analyses suggest that GWAS will continue to be highly productive for the discovery of additional susceptibility loci for common diseases.     

Recombination and linkage disequilibrium in Arabidopsis thaliana

Linkage disequilibrium (LD) is a major aspect of the organization of genetic variation in natural populations. Here we describe the genome-wide pattern of LD in a sample of 19 Arabidopsis thaliana accessions using 341,602 non-singleton SNPs. LD decays within 10 kb on average, considerably faster than previously estimated. Tag SNP selection algorithms and 'hide-the-SNP' simulations suggest that genome-wide association mapping will require only 40%-50% of the observed SNPs, a reduction similar to estimates in a sample of African Americans. An Affymetrix genotyping array containing 250,000 SNPs has been designed based on these results; we demonstrate that it should have more than adequate coverage for genome-wide association mapping. The extent of LD is highly variable, and we find clear evidence of recombination hotspots, which seem to occur preferentially in intergenic regions. LD also reflects the action of selection, and it is more extensive between nonsynonymous polymorphisms than between synonymous polymorphisms.     

A new multipoint method for genome-wide association studies by imputation of genotypes

Genome-wide association studies are set to become the method of choice for uncovering the genetic basis of human diseases. A central challenge in this area is the development of powerful multipoint methods that can detect causal variants that have not been directly genotyped. We propose a coherent analysis framework that treats the problem as one involving missing or uncertain genotypes. Central to our approach is a model-based imputation method for inferring genotypes at observed or unobserved SNPs, leading to improved power over existing methods for multipoint association mapping. Using real genome-wide association study data, we show that our approach (i) is accurate and well calibrated, (ii) provides detailed views of associated regions that facilitate follow-up studies and (iii) can be used to validate and correct data at genotyped markers. A notable future use of our method will be to boost power by combining data from genome-wide scans that use different SNP sets.     

The Arabidopsis lyrata genome sequence and the basis of rapid genome size change

We report the 207-Mb genome sequence of the North American Arabidopsis lyrata strain MN47 based on 8.3× dideoxy sequence coverage. We predict 32,670 genes in this outcrossing species compared to the 27,025 genes in the selfing species Arabidopsis thaliana. The much smaller 125-Mb genome of A. thaliana, which diverged from A. lyrata 10 million years ago, likely constitutes the derived state for the family. We found evidence for DNA loss from large-scale rearrangements, but most of the difference in genome size can be attributed to hundreds of thousands of small deletions, mostly in noncoding DNA and transposons. Analysis of deletions and insertions still segregating in A. thaliana indicates that the process of DNA loss is ongoing, suggesting pervasive selection for a smaller genome. The high-quality reference genome sequence for A. lyrata will be an important resource for functional, evolutionary and ecological studies in the genus Arabidopsis.     

Screening a large reference sample to identify very low frequency sequence variants: comparisons between two genes

Most human sequence variation is in the form of single-nucleotide polymorphisms (SNPs). It has been proposed that coding-region SNPs (cSNPs) be used for direct association studies to determine the genetic basis of complex traits. The success of such studies depends on the frequency of disease-associated alleles, and their distribution in different ethnic populations. If disease-associated alleles are frequent in most populations, then direct genotyping of candidate variants could show robust associations in manageable study samples. This approach is less feasible if the genetic risk from a given candidate gene is due to many infrequent alleles. Previous studies of several genes demonstrated that most variants are relatively infrequent (<0.05). These surveys genotyped small samples (n<75) and thus had limited ability to identify rare alleles. Here we evaluate the prevalence and distribution of such rare alleles by genotyping an ethnically diverse reference sample that is more than six times larger than those used in previous studies (n=450). We screened for variants in the complete coding sequence and intron-exon junctions of two candidate genes for neuropsychiatric phenotypes: SLC6A4, encoding the serotonin transporter; and SLC18A2, encoding the vesicular monoamine transporter. Both genes have unique roles in neuronal transmission, and variants in either gene might be associated with neurobehavioral phenotypes.     

Common variants on chromosomes 2q35 and 16q12 confer susceptibility to estrogen receptor-positive breast cancer

Familial clustering studies indicate that breast cancer risk has a substantial genetic component. To identify new breast cancer risk variants, we genotyped approximately 300,000 SNPs in 1,600 Icelandic individuals with breast cancer and 11,563 controls using the Illumina Hap300 platform. We then tested selected SNPs in five replication sample sets. Overall, we studied 4,554 affected individuals and 17,577 controls. Two SNPs consistently associated with breast cancer: approximately 25% of individuals of European descent are homozygous for allele A of rs13387042 on chromosome 2q35 and have an estimated 1.44-fold greater risk than noncarriers, and for allele T of rs3803662 on 16q12, about 7% are homozygous and have a 1.64-fold greater risk. Risk from both alleles was confined to estrogen receptor-positive tumors. At present, no genes have been identified in the linkage disequilibrium block containing rs13387042. rs3803662 is near the 5' end of TNRC9 , a high mobility group chromatin-associated protein whose expression is implicated in breast cancer metastasis to bone.     

Resequencing of 31 wild and cultivated soybean genomes identifies patterns of genetic diversity and selection

We report a large-scale analysis of the patterns of genome-wide genetic variation in soybeans. We re-sequenced a total of 17 wild and 14 cultivated soybean genomes to an average of approximately ×5 depth and >90% coverage using the Illumina Genome Analyzer II platform. We compared the patterns of genetic variation between wild and cultivated soybeans and identified higher allelic diversity in wild soybeans. We identified a high level of linkage disequilibrium in the soybean genome, suggesting that marker-assisted breeding of soybean will be less challenging than map-based cloning. We report linkage disequilibrium block location and distribution, and we identified a set of 205,614 tag SNPs that may be useful for QTL mapping and association studies. The data here provide a valuable resource for the analysis of wild soybeans and to facilitate future breeding and quantitative trait analysis.     

Characterization of single-nucleotide polymorphisms in coding regions of human genes.

A major goal in human genetics is to understand the role of common genetic variants in susceptibility to common diseases. This will require characterizing the nature of gene variation in human populations, assembling an extensive catalogue of single-nucleotide polymorphisms (SNPs) in candidate genes and performing association studies for particular diseases. At present, our knowledge of human gene variation remains rudimentary. Here we describe a systematic survey of SNPs in the coding regions of human genes. We identified SNPs in 106 genes relevant to cardiovascular disease, endocrinology and neuropsychiatry by screening an average of 114 independent alleles using 2 independent screening methods. To ensure high accuracy, all reported SNPs were confirmed by DNA sequencing. We identified 560 SNPs, including 392 coding-region SNPs (cSNPs) divided roughly equally between those causing synonymous and non-synonymous changes. We observed different rates of polymorphism among classes of sites within genes (non-coding, degenerate and non-degenerate) as well as between genes. The cSNPs most likely to influence disease, those that alter the amino acid sequence of the encoded protein, are found at a lower rate and with lower allele frequencies than silent substitutions. This likely reflects selection acting against deleterious alleles during human evolution. The lower allele frequency of missense cSNPs has implications for the compilation of a comprehensive catalogue, as well as for the subsequent application to disease association.     

Genome partitioning of genetic variation for complex traits using common SNPs

We estimate and partition genetic variation for height, body mass index (BMI), von Willebrand factor and QT interval (QTi) using 586,898 SNPs genotyped on 11,586 unrelated individuals. We estimate that ～45%, ～17%, ～25% and ～21% of the variance in height, BMI, von Willebrand factor and QTi, respectively, can be explained by all autosomal SNPs and a further ～0.5-1% can be explained by X chromosome SNPs. We show that the variance explained by each chromosome is proportional to its length, and that SNPs in or near genes explain more variation than SNPs between genes. We propose a new approach to estimate variation due to cryptic relatedness and population stratification. Our results provide further evidence that a substantial proportion of heritability is captured by common SNPs, that height, BMI and QTi are highly polygenic traits, and that the additive variation explained by a part of the genome is approximately proportional to the total length of DNA contained within genes therein.     

Additional SNPs and linkage-disequilibrium analyses are necessary for whole-genome association studies in humans

More than 5 million single-nucleotide polymorphisms (SNPs) with minor-allele frequency greater than 10% are expected to exist in the human genome. Some of these SNPs may be associated with risk of developing common diseases. To assess the power of currently available SNPs to detect such associations, we resequenced 50 genes in two ethnic samples and measured patterns of linkage disequilibrium between the subset of SNPs reported in dbSNP and the complete set of common SNPs. Our results suggest that using all 2.7 million SNPs currently in the database would detect nearly 80% of all common SNPs in European populations but only 50% of those common in the African American population and that efficient selection of a minimal subset of SNPs for use in association studies requires measurement of allele frequency and linkage disequilibrium relationships for all SNPs in dbSNP.     

Estimating the proportion of variation in susceptibility to schizophrenia captured by common SNPs

Schizophrenia is a complex disorder caused by both genetic and environmental factors. Using 9,087 affected individuals, 12,171 controls and 915,354 imputed SNPs from the Schizophrenia Psychiatric Genome-Wide Association Study (GWAS) Consortium (PGC-SCZ), we estimate that 23% (s.e. = 1%) of variation in liability to schizophrenia is captured by SNPs. We show that a substantial proportion of this variation must be the result of common causal variants, that the variance explained by each chromosome is linearly related to its length (r = 0.89, P = 2.6 × 10(-8)), that the genetic basis of schizophrenia is the same in males and females, and that a disproportionate proportion of variation is attributable to a set of 2,725 genes expressed in the central nervous system (CNS; P = 7.6 × 10(-8)). These results are consistent with a polygenic genetic architecture and imply more individual SNP associations will be detected for this disease as sample size increases.     

Large-scale genome-wide association analysis of bipolar disorder identifies a new susceptibility locus near ODZ4

We conducted a combined genome-wide association study (GWAS) of 7,481 individuals with bipolar disorder (cases) and 9,250 controls as part of the Psychiatric GWAS Consortium. Our replication study tested 34 SNPs in 4,496 independent cases with bipolar disorder and 42,422 independent controls and found that 18 of 34 SNPs had P < 0.05, with 31 of 34 SNPs having signals with the same direction of effect (P = 3.8 × 10(-7)). An analysis of all 11,974 bipolar disorder cases and 51,792 controls confirmed genome-wide significant evidence of association for CACNA1C and identified a new intronic variant in ODZ4. We identified a pathway comprised of subunits of calcium channels enriched in bipolar disorder association intervals. Finally, a combined GWAS analysis of schizophrenia and bipolar disorder yielded strong association evidence for SNPs in CACNA1C and in the region of NEK4-ITIH1-ITIH3-ITIH4. Our replication results imply that increasing sample sizes in bipolar disorder will confirm many additional loci.     

Genome assembly comparison identifies structural variants in the human genome

Numerous types of DNA variation exist, ranging from SNPs to larger structural alterations such as copy number variants (CNVs) and inversions. Alignment of DNA sequence from different sources has been used to identify SNPs and intermediate-sized variants (ISVs). However, only a small proportion of total heterogeneity is characterized, and little is known of the characteristics of most smaller-sized (<50 kb) variants. Here we show that genome assembly comparison is a robust approach for identification of all classes of genetic variation. Through comparison of two human assemblies (Celera's R27c compilation and the Build 35 reference sequence), we identified megabases of sequence (in the form of 13,534 putative non-SNP events) that were absent, inverted or polymorphic in one assembly. Database comparison and laboratory experimentation further demonstrated overlap or validation for 240 variable regions and confirmed >1.5 million SNPs. Some differences were simple insertions and deletions, but in regions containing CNVs, segmental duplication and repetitive DNA, they were more complex. Our results uncover substantial undescribed variation in humans, highlighting the need for comprehensive annotation strategies to fully interpret genome scanning and personalized sequencing projects.