Imputed Single Nucleotide Polymorphisms Research Articles

Resequencing and Analysis of Variation in the TCF7L2 Gene in African Americans Suggests That SNP rs7903146 Is the Causal Diabetes Susceptibility Variant

OBJECTIVEVariation in the transcription factor 7-like 2 (TCF7L2) locus is associated with type 2 diabetes across multiple ethnicities. The aim of this study was to elucidate which variant in TCF7L2 confers diabetes susceptibility in African Americans.RESEARCH DESIGN AND METHODSThrough the evaluation of tagging single nucleotide polymorphisms (SNPs), type 2 diabetes susceptibility was limited to a 4.3-kb interval, which contains the YRI (African) linkage disequilibrium (LD) block containing rs7903146. To better define the relationship between type 2 diabetes risk and genetic variation we resequenced this 4.3-kb region in 96 African American DNAs. Thirty-three novel and 13 known SNPs were identified: 20 with minor allele frequencies (MAF) >0.05 and 12 with MAF >0.10. These polymorphisms and the previously identified DG10S478 microsatellite were evaluated in African American type 2 diabetic cases (n = 1,033) and controls (n = 1,106).RESULTSVariants identified from direct sequencing and databases were genotyped or imputed. Fifteen SNPs showed association with type 2 diabetes (P < 0.05) with rs7903146 being the most significant (P = 6.32 × 10−6). Results of imputation, haplotype, and conditional analysis of SNPs were consistent with rs7903146 being the trait-defining SNP. Analysis of the DG10S478 microsatellite, which is outside the 4.3-kb LD block, revealed consistent association of risk allele 8 with type 2 diabetes (odds ratio [OR] = 1.33; P = 0.022) as reported in European populations; however, allele 16 (MAF = 0.016 cases and 0.032 controls) was strongly associated with reduced risk (OR = 0.39; P = 5.02 × 10−5) in contrast with previous studies.CONCLUSIONSIn African Americans, these observations suggest that rs7903146 is the trait-defining polymorphism associated with type 2 diabetes risk. Collectively, these results support ethnic differences in type 2 diabetes associations.

Family‐based association analysis to finemap bipolar linkage peak on chromosome 8q24 using 2,500 genotyped SNPs and 15,000 imputed SNPs

Multiple linkage and association studies have suggested chromosome 8q24 as a promising candidate region for bipolar disorder (BP). We performed a detailed association analysis assessing the contribution of common genetic variation in this region to the risk of BP. We analyzed 2,756 single nucleotide polymorphism (SNP) markers in the chromosome 8q24 region of 3,512 individuals from 737 families. In addition, we extended genotype imputation methods to family-based data and imputed 22,725 HapMap SNPs in the same region on 8q24. We applied a family-based method to test 15,552 high-quality genotyped or imputed SNPs for association with BP. Our association analysis identified the most significant marker (p=4.80 × 10(-5) ), near the gene encoding potassium voltage-gated channel KQT-like protein (KCNQ3). Other marginally significant markers were located near adenylate cyclase 8 (ADCY8) and ST3 beta-galactoside alpha-2,3-sialyltransferase 1 (ST3GAL1). We developed an approach to apply MACH imputation to family-based data, which can increase the power to detect association signals. Our association results showed suggestive evidence of association of BP with loci near KCNQ3, ADCY8, and ST3GAL1. Consistent with genes identified by genome-wide association studies for BP, our results suggest the involvement of ion channelopathy in BP pathogenesis. However, common variants are insufficient to explain linkage findings in 8q24; other genetic variation should be explored.

A Meta-Analysis of Nodular Sclerosis Hodgkin Lymphoma Identifies Risk Loci at 6p21.32

AbstractAbstract 2686Nodular sclerosis Hodgkin lymphoma (NSHL) is epidemiologically and pathologically distinct from other Hodgkin lymphoma subtypes. It is highly heritable and is the most common form of HL in young adults. In a combined analysis of results from two genome-wide association scans (one undertaken at the University of Southern California using the Illumina Human610-Quad beadchip and the other at the University of Chicago using the Affymetrix Genome-Wide Human SNP Array 6.0) of 395 European-origin NSHL patients and 3,245 controls, we identified two single nucleotide polymorphisms (SNPs) at chromosome 6p21.32 directly genotyped in both scans associated with disease risk at genome-wide levels of significance (rs9268542: p=8.2×10−10, odds ratio (OR)= 1.63, 95% Confidence Interval (CI) = 1.40–1.91; and rs204999: p=2.3×10-9, OR= 0.54, 95% CI: 0.44=0.66). These SNPs are located in distinct haplotype blocks and are in only weak linkage disequilibrium (LD) (r2=0.06). Joint analysis revealed that each SNP was independently associated with disease risk, with the global likelihood ratio test remaining statistically significant for each group (p-value for 2 df test USC = 7.46 × 10-9 and UC=.3.51 × 10-6). Neither SNP is in a known gene or microRNA. rs9268542 is in LD with SNPs in the HLA-DRA gene (r2=0.92-0.93) and is a cis-acting expression QTL (eQTL) for several HLA Class II genes (HLA-DRB1-HLA-DRB5; HLA-DQB1-HLA-DQB2; all p=1×10-5). Of note, rs6903608, an imputed SNP in the same region also surpassing genome-wide significance for association with NSHL, is an eQTL for these same HLA Class II genes at p=1×10-20. rs204999 is in LD with SNPs in the nearby PRRT1 (r2=0.82) and EGFL8 (r2=0.81) genes, for which function remains poorly defined. Thus, we identified one SNP on chromosome 6p21.32 associated with NSHL that, by eQTL analysis, may be implicated in the transcriptional regulation of HLA-class II genes, a known susceptibly region for NSHL, and one SNP in the genomic region that appears to be independently-associated with NSHL, but which does not appear to be linked to the HLA locus. Disclosures:No relevant conflicts of interest to declare.

Genome-Wide Association Study Identifies Multiple Genetic Loci for Activated Partial Thromboplastin Time and Prothrombin Time

AbstractAbstract 4222 Background:Activated partial thromboplastin time (aPTT) and prothrombin time (PT) are commonly used to screen for coagulation factor deficiencies. Shorter aPTT is also a risk marker for incident and recurrent venous thromboembolism (VTE). Genetic factors influencing aPTT and PT are not well understood. So far only one genome-wide association study (GWAS) has been reported for aPTT in 1,477 participants and none for PT. Methods:We conducted a GWAS for the aPTT in 9,240 European Americans (EAs) from the Atherosclerosis Risk in Communities (ARIC) study and for PT in 1,221 EAs from the Genetic Study of Three Population Microisolates in South Tyrol (MICROS). Replication was assessed by in silico analysis in MICROS (aPTT, n=1,215) and the Lothian Birth Cohorts (LBC) (LBC1936 (aPTT and PT, n=925-989), LBC1921 (aPTT, n=445)), and by de novo genotyping in the Caerphilly study (aPTT, n=882). Subjects on anticoagulants were excluded. Genotyping was conducted with the Affymetrix single nucleotide polymorphism (SNP) array 6.0 or Illumina HumanHap300/370 and imputed to ∼2.5 million HapMap SNPs. SNPs with imputation quality score < 0.3 or minor allele frequencies ≤1% were excluded from data analysis. The imputed SNP dosages were analyzed in linear regression adjusted for age, sex, and field center, where applicable. Results:Five loci were associated with aPTT at genome-wide significance of p<5×10-8 that have not been previously reported: F5 (1q23, top SNP rs9332701, missense, β (effect size associated with one copy increase in minor allele)=0.54, p=3.7×10-8), F11 (4q35, rs1593, intronic, β =0.54 and 0.36, p=1.25×10-17 and 2.0×10-8 before and after adjusting for known variants in F11), NSD1 (5q35.2-q35.3, rs11950938, intronic, β =1.00 and 1.21, p=1.11×10-15 and 1.5×10-22 before and after adjusting for known variants in F12 of the same region), C6orf10 (6p21.3, rs2050190, intronic, β =-0.25, p=1.3×10-8), and ABO (9q34.1-q34.2, rs8176704, intronic, tag for A2 group, β =0.19 and 0.89, p=0.02 and 4.3×10-24 before and after adjusting for O blood group). Three of the five loci replicated in at least one replication sample and the other two were directionally consistent in 3 replication samples. Furthermore, meta-analysis pooling the discovery and replication GWAS samples yielded two additional independent loci at chromosomes 1q23 (F5, best SNP rs6028, coding-synonymous, β =0.23, p=5.97×10-9) and 15q25.3 (AGBL1, rs2469184, intronic, β =0.16, p= 4.24×10-8), with consistent associations across studies. The signals at the F5 region were not due to FV Leiden (rs6025, p=0.46 in ARIC). We also confirmed previously reported loci in KNG1, HRG, F11, F12, and ABO (O group). For PT, novel associations from two gene regions reached genome-wide significance in MICROS: F7 (top SNP rs3093253, within an exon but not translated, β =-5.44, p=8.2×10-19) and PROCR (rs6060244, near PROCR, β =4.11, p=3.5×10-9). Both loci replicated in LBC1936. Conclusions:In this large GWAS, six of the nine novel loci associated with the aPTT and PT are coagulation-related and the other three (NSD1, C6orf10, and AGBL1) are new candidate genes not directly involved in coagulation. The C6orf10 gene interacts with TNF-a at the transcription level and was previously associated with inflammatory diseases. These findings may be relevant to the prevention and treatment of coagulation disorders including VTE. Disclosures:No relevant conflicts of interest to declare.

Genome-wide association study of major depressive disorder: new results, meta-analysis, and lessons learned.

Major depressive disorder (MDD) is a common complex disorder with a partly genetic etiology. We conducted a genome-wide association study of the MDD2000+ sample (2431 cases, 3673 screened controls and >1 M imputed single-nucleotide polymorphisms (SNPs)). No SNPs achieved genome-wide significance either in the MDD2000+ study, or in meta-analysis with two other studies totaling 5763 cases and 6901 controls. These results imply that common variants of intermediate or large effect do not have main effects in the genetic architecture of MDD. Suggestive but notable results were (a) gene-based tests suggesting roles for adenylate cyclase 3 (ADCY3, 2p23.3) and galanin (GAL, 11q13.3); published functional evidence relates both of these to MDD and serotonergic signaling; (b) support for the bipolar disorder risk variant SNP rs1006737 in CACNA1C (P=0.020, odds ratio=1.10); and (c) lack of support for rs2251219, a SNP identified in a meta-analysis of affective disorder studies (P=0.51). We estimate that sample sizes 1.8- to 2.4-fold greater are needed for association studies of MDD compared with those for schizophrenia to detect variants that explain the same proportion of total variance in liability. Larger study cohorts characterized for genetic and environmental risk factors accumulated prospectively are likely to be needed to dissect more fully the etiology of MDD.

Germline Variation in Apoptosis Pathway Genes and Risk of Non–Hodgkin's Lymphoma

The t(14;18)(q32;q21) translocation is the most commonly observed chromosomal translocation in non-Hodgkin's lymphoma (NHL), resulting in constitutive Bcl-2 expression and apoptosis inhibition. In addition, germline variation in both BCL2L11 (BIM) and CASP9, known regulators of apoptosis, has recently been linked to NHL risk. We conducted a comprehensive evaluation of 36 apoptosis pathway genes with risk of NHL. We genotyped 226 single-nucleotide polymorphisms (SNP) from 36 candidate genes in a clinic-based study of 441 newly diagnosed NHL cases and 475 frequency-matched controls. We used principal components analysis to assess gene-level associations, and logistic regression to assess SNP-level associations. MACH was used for imputation of SNPs in BCL2L11 and CASP9. In gene-level analyses, BCL2L11 (P = 0.0019), BCLAF1 (P = 0.0097), BAG5 (P = 0.026), and CASP9 (P = 0.0022) were associated with NHL risk after accounting for multiple testing (tail strength, 0.38; 95% confidence interval, 0.05-0.70). Two of the five BCL2L11 tagSNPs (rs6746608 and rs12613243), both genotyped BCLAF1 tagSNPs (rs797558 and rs703193), the single genotyped BAG5 tagSNP (rs7693), and three of the seven genotyped CASP9 tagSNPs (rs6685648, rs2020902, and rs2042370) were significant at P < 0.05. We successfully imputed BCL2L11 and CASP9 SNPs previously linked to NHL, and replicated all four BCL2L11 and two of three CASP9 SNPs. We replicated the association of BCL2L11 and CASP9 with NHL risk at the gene and SNP level, and identified novel associations with BCLAF1 and BAG5. Closer evaluation of germline variation of genes in the apoptosis pathway with risk of NHL and its subtypes is warranted.

Accuracy of direct genomic values derived from imputed single nucleotide polymorphism genotypes in Jersey cattle

The objective of the present study was to evaluate the predictive ability of direct genomic values for economically important dairy traits when genotypes at some single nucleotide polymorphism (SNP) loci were imputed rather than measured directly. Genotypic data consisted of 42,552 SNP genotypes for each of 1,762 Jersey sires. Phenotypic data consisted of predicted transmitting abilities (PTA) for milk yield, protein percentage, and daughter pregnancy rate from May 2006 for 1,446 sires in the training set and from April 2009 for 316 sires in the testing set. The SNP effects were estimated using the Bayesian least absolute selection and shrinkage operator (LASSO) method with data of sires in the training set, and direct genomic values (DGV) for sires in the testing set were computed by multiplying these estimates by corresponding genotype dosages for sires in the testing set. The mean correlation across traits between DGV (before progeny testing) and PTA (after progeny testing) for sires in the testing set was 70.6% when all 42,552 SNP genotypes were used. When genotypes for 93.1, 96.6, 98.3, or 99.1% of loci were masked and subsequently imputed in the testing set, mean correlations across traits between DGV and PTA were 68.5, 64.8, 54.8, or 43.5%, respectively. When genotypes were also masked and imputed for a random 50% of sires in the training set, mean correlations across traits between DGV and PTA were 65.7, 63.2, 53.9, or 49.5%, respectively. Results of this study indicate that if a suitable reference population with high-density genotypes is available, a low-density chip comprising 3,000 equally spaced SNP may provide approximately 95% of the predictive ability observed with the BovineSNP50 Beadchip (Illumina Inc., San Diego, CA) in Jersey cattle. However, if fewer than 1,500 SNP are genotyped, the accuracy of DGV may be limited by errors in the imputed genotypes of selection candidates.

Bayesian epistasis association mapping via SNP imputation

Genetic mutations may interact to increase the risk of human complex diseases. Mapping of multiple interacting disease loci in the human genome has recently shown promise in detecting genes with little main effects. The power of interaction association mapping, however, can be greatly influenced by the set of single nucleotide polymorphism (SNP) genotyped in a case-control study. Previous imputation methods only focus on imputation of individual SNPs without considering their joint distribution of possible interactions. We present a new method that simultaneously detects multilocus interaction associations and imputes missing SNPs from a full Bayesian model. Our method treats both the case-control sample and the reference data as random observations. The output of our method is the posterior probabilities of SNPs for their marginal and interacting associations with the disease. Using simulations, we show that the method produces accurate and robust imputation with little overfitting problems. We further show that, with the type I error rate maintained at a common level, SNP imputation can consistently and sometimes substantially improve the power of detecting disease interaction associations. We use a data set of inflammatory bowel disease to demonstrate the application of our method.

Founder population-specific HapMap panel increases power in GWA studies through improved imputation accuracy and CNV tagging

The combining of genome-wide association (GWA) data across populations represents a major challenge for massive global meta-analyses. Genotype imputation using densely genotyped reference samples facilitates the combination of data across different genotyping platforms. HapMap data is typically used as a reference for single nucleotide polymorphism (SNP) imputation and tagging copy number polymorphisms (CNPs). However, the advantage of having population-specific reference panels for founder populations has not been evaluated. We looked at the properties and impact of adding 81 individuals from a founder population to HapMap3 reference data on imputation quality, CNP tagging, and power to detect association in simulations and in an independent cohort of 2138 individuals. The gain in SNP imputation accuracy was highest among low-frequency markers (minor allele frequency [MAF] < 5%), for which adding the population-specific samples to the reference set increased the median R(2) between imputed and genotyped SNPs from 0.90 to 0.94. Accuracy also increased in regions with high recombination rates. Similarly, a reference set with population-specific extension facilitated the identification of better tag-SNPs for a subset of CNPs; for 4% of CNPs the R(2) between SNP genotypes and CNP intensity in the independent population cohort was at least twice as high as without the extension. We conclude that even a relatively small population-specific reference set yields considerable benefits in SNP imputation, CNP tagging accuracy, and the power to detect associations in founder populations and population isolates in particular.

Imputation aware meta-analysis of genome-wide association studies

Genome-wide association studies have recently identified many new loci associated with human complex diseases. These newly discovered variants typically have weak effects requiring studies with large numbers of individuals to achieve the statistical power necessary to identify them. Likely, there exist even more associated variants, which remain to be found if even larger association studies can be assembled. Meta-analysis provides a straightforward means of increasing study sample sizes without collecting new samples by combining existing data sets. One obstacle to combining studies is that they are often performed on platforms with different marker sets. Current studies overcome this issue by imputing genotypes missing from each of the studies and then performing standard meta-analysis techniques. We show that this approach may result in a loss of power since errors in imputation are not accounted for. We present a new method for performing meta-analysis over imputed single nucleotide polymorphisms, show that it is optimal with respect to power, and discuss practical implementation issues. Through simulation experiments, we show that our imputation aware meta-analysis approach outperforms or matches standard meta-analysis approaches.

ProbABEL package for genome-wide association analysis of imputed data

BackgroundOver the last few years, genome-wide association (GWA) studies became a tool of choice for the identification of loci associated with complex traits. Currently, imputed single nucleotide polymorphisms (SNP) data are frequently used in GWA analyzes. Correct analysis of imputed data calls for the implementation of specific methods which take genotype imputation uncertainty into account.ResultsWe developed the ProbABEL software package for the analysis of genome-wide imputed SNP data and quantitative, binary, and time-till-event outcomes under linear, logistic, and Cox proportional hazards models, respectively. For quantitative traits, the package also implements a fast two-step mixed model-based score test for association in samples with differential relationships, facilitating analysis in family-based studies, studies performed in human genetically isolated populations and outbred animal populations.ConclusionsProbABEL package provides fast efficient way to analyze imputed data in genome-wide context and will facilitate future identification of complex trait loci.

Genetic analysis of albuminuria in aging mice and concordance with loci for human diabetic nephropathy found in a genome-wide association scan

Aging in the kidney can cause albuminuria, and discovering molecular mechanisms responsible for this might offer a new perspective on the etiology of this abnormality. Haplotype association mapping in the mouse is a novel approach which uses the haplotypes of the relatively closely related mouse inbred strains and the phenotypic variation among these strains to find associations between haplotypes and phenotype. The albumin-to-creatinine ratios, a measure of urinary albumin excretion, were determined in 30 inbred mouse strains at 12, 18, and 24 months of age. Mapping was performed for males and females separately at all three time points using a high density set of 63,222 single-nucleotide polymorphisms to determine genetic loci involved in albuminuria. One significant and eight suggestive loci were found, some of which map to previously identified loci for traits associated with kidney damage in the mouse, but with a much higher resolution thus narrowing their chromosomal location. These nine loci were then compared with genome-wide association scans for diabetic nephropathy (DN) in human type I diabetes. Our study found that two of the nine mouse loci for age-related albuminuria were significantly associated with DN and consistent across male and female strata. This suggests common underlying genes predispose to kidney disease in mice and humans.

CGDSNPdb: a database resource for error-checked and imputed mouse SNPs

The Center for Genome Dynamics Single Nucleotide Polymorphism Database (CGDSNPdb) is an open-source value-added database with more than nine million mouse single nucleotide polymorphisms (SNPs), drawn from multiple sources, with genotypes assigned to multiple inbred strains of laboratory mice. All SNPs are checked for accuracy and annotated for properties specific to the SNP as well as those implied by changes to overlapping protein-coding genes. CGDSNPdb serves as the primary interface to two unique data sets, the ‘imputed genotype resource’ in which a Hidden Markov Model was used to assess local haplotypes and the most probable base assignment at several million genomic loci in tens of strains of mice, and the Affymetrix Mouse Diversity Genotyping Array, a high density microarray with over 600 000 SNPs and over 900 000 invariant genomic probes. CGDSNPdb is accessible online through either a web-based query tool or a MySQL public login.Database URL: http://cgd.jax.org/cgdsnpdb/

Association study of bromodomain‐containing 1 gene with schizophrenia in Japanese population

Chromosome 22q13 region has been implicated in schizophrenia in several linkage studies. Genes within this locus are therefore promising genetic and biologic candidate genes for schizophrenia if they are expressed in the brain or predicted to have some role in brain development. A recent study reported that bromodomain-containing 1 gene (BRD1), located in 22q13, showed an association with schizophrenia in a Scottish population. Except for being a putative regulator of transcription, the precise function of BRD1 is not clear; however, expression analysis of BRD1 mRNA revealed widespread expression in mammalian brains. In our study, we explored the association of BRD1 with schizophrenia in a Japanese population (626 cases and 770 controls). In this association analysis, we first examined 10 directly genotyped single-nucleotide polymorphisms (SNPs) and 20 imputed SNPs. Second, we compared the BRD1 mRNA levels between cases and controls using lymphoblastoid cell lines (LCLs) derived from 29 cases and 30 controls. Although the SNP (rs138880) that previously has been associated with schizophrenia showed the same trend in the Japanese population, no significant association was detected between BRD1 and schizophrenia in our study. Similarly, no significant differences in BRD1 mRNA levels in LCLs were detected. Taken together, we could not strongly show that common SNPs in the BRD1 gene account for a substantial proportion of the genetic risk for schizophrenia in the Japanese population.

The dopamine D3 receptor ( DRD3) gene and risk of schizophrenia: Case–control studies and an updated meta-analysis

The dopamine D3 receptor (DRD3) has been suggested to be involved in the pathophysiology of schizophrenia. DRD3 has been tested for an association with schizophrenia, but with conflicting results. A recent meta-analysis suggested that the haplotype T–T–T–G for the SNPs rs7631540–rs1486012–rs2134655–rs963468 may confer protection against schizophrenia. However, almost all previous studies of the association between DRD3 and schizophrenia have been performed using a relatively small sample size and a limited number of markers. To assess whether DRD3 is implicated in vulnerability to schizophrenia, we conducted case–control association studies and performed an updated meta-analysis. In the first population (595 patients and 598 controls), we examined 16 genotyped single nucleotide polymorphisms (SNPs), including tagging SNPs selected from the HapMap database and SNPs detected through resequencing, as well as 58 imputed SNPs that are not directly genotyped. To confirm the results obtained, we genotyped the SNPs rs7631540–rs1486012–rs2134655–rs963468 in a second, independent population (2126 patients and 2228 controls). We also performed an updated meta-analysis of the haplotype, combining the results obtained in five populations, with a total sample size of 7551. No supportive evidence was obtained for an association between DRD3 and schizophrenia in our Japanese subjects. Our updated meta-analysis also failed to confirm the existence of a protective haplotype. To draw a definitive conclusion, further studies using larger samples and sufficient markers should be carried out in various ethnic populations.

Genome‐wide associations studies for melanoma and nevi

Genome‐wide associations studies for melanoma and nevi

Identification of Las2, a Major Modifier Gene Affecting the Pas1 Mouse Lung Tumor Susceptibility Locus

Lung cancer is the leading cause of cancer death worldwide. Here, we describe a genome-wide association study of chemically induced lung tumorigenesis on 593 mice from 21 inbred strains using 115,904 genotyped and 1,952,918 imputed single nucleotide polymorphisms (SNPs). Using a genetic background-controlled genome search, we identified a novel lung tumor susceptibility gene Las2 (Lung adenoma susceptibility 2) on distal chromosome 18. Las2 showed strong association with resistance to tumor induction (rs30245983; P = 1.87 x 10(-9)) as well as epistatic interactions (P = 1.71 x 10(-3)) with the pulmonary adenoma susceptibility 1 locus, a major locus affecting mouse lung tumor development (rs13459098, P = 5.64 x 10(-27)). Sequencing analysis revealed four nonsynonymous SNPs and two insertions/deletions in the susceptible allele of Las2, resulting in the loss of tumor suppressor activities in both cell colony formation and nude mouse tumorigenicity assays. Deletion of LAS2 was observed in approximately 40% of human lung adenocarcinomas, implying that loss of function of LAS2 may be a key step for lung tumorigenesis.

SNP imputation in association studies

Only a subset of single-nucleotide polymorphisms (SNPs) can be genotyped in genome-wide association studies. Imputation methods can infer the alleles of 'hidden' variants and use those inferences to test the hidden variants for association.

A Genome-Wide Association Study Confirms VKORC1, CYP2C9, and CYP4F2 as Principal Genetic Determinants of Warfarin Dose

We report the first genome-wide association study (GWAS) whose sample size (1,053 Swedish subjects) is sufficiently powered to detect genome-wide significance (p<1.5×10−7) for polymorphisms that modestly alter therapeutic warfarin dose. The anticoagulant drug warfarin is widely prescribed for reducing the risk of stroke, thrombosis, pulmonary embolism, and coronary malfunction. However, Caucasians vary widely (20-fold) in the dose needed for therapeutic anticoagulation, and hence prescribed doses may be too low (risking serious illness) or too high (risking severe bleeding). Prior work established that ∼30% of the dose variance is explained by single nucleotide polymorphisms (SNPs) in the warfarin drug target VKORC1 and another ∼12% by two non-synonymous SNPs (*2, *3) in the cytochrome P450 warfarin-metabolizing gene CYP2C9. We initially tested each of 325,997 GWAS SNPs for association with warfarin dose by univariate regression and found the strongest statistical signals (p<10−78) at SNPs clustering near VKORC1 and the second lowest p-values (p<10−31) emanating from CYP2C9. No other SNPs approached genome-wide significance. To enhance detection of weaker effects, we conducted multiple regression adjusting for known influences on warfarin dose (VKORC1, CYP2C9, age, gender) and identified a single SNP (rs2108622) with genome-wide significance (p = 8.3×10−10) that alters protein coding of the CYP4F2 gene. We confirmed this result in 588 additional Swedish patients (p<0.0029) and, during our investigation, a second group provided independent confirmation from a scan of warfarin-metabolizing genes. We also thoroughly investigated copy number variations, haplotypes, and imputed SNPs, but found no additional highly significant warfarin associations. We present power analysis of our GWAS that is generalizable to other studies, and conclude we had 80% power to detect genome-wide significance for common causative variants or markers explaining at least 1.5% of dose variance. These GWAS results provide further impetus for conducting large-scale trials assessing patient benefit from genotype-based forecasting of warfarin dose.

Pharmacological effects of lipid-lowering drugs recapitulate with a larger amplitude the phenotypic effects of common variants within their target genes

A major expectation underlying the search for novel susceptibility genes for common diseases using genome-wide association studies (GWAS) is that these discoveries will lead to new drug targets. This claim has not been verified yet. Here, we tested the hypothesis that common single nucleotide polymorphisms (SNPs) within drug target genes are associated with the corresponding phenotypes, using a population-based GWAS dataset and lipid-lowering drugs as a test case. We examined the association between 36 genotyped and 193 imputed SNPs within four lipid-lowering drug target genes (HMGCR, PPARA, HM74A/GPR109A and CETP) and four non-lipid drug target genes (ACE, AGTR1, P2RY12, and ATP4B) and lipid phenotypes, blood pressure, and coronary artery disease in 5635 adult participants of the Lausanne, Switzerland, CoLaus study, genotyped using the Affymetrix 500K SNP chip technology. The phenotypes associated with SNPs within drug target genes recapitulated to a certain extent the pharmacological effects of the drug. The amplitude of the SNP effect was about 10 times smaller than the pharmacological effect of the corresponding drug. In particular, several CETP SNPs were associated with an elevation in HDL-cholesterol levels, yet a lower diastolic blood pressure, providing evidence that the blood pressure elevation induced by the CETP inhibitor torcetrapib is more likely compound specific than class specific. Pharmacological modulation of lipid-lowering drug targets recapitulates, and markedly amplifies, the phenotypic effects of common SNPs within these target genes. This data provides indirect evidence that, with certain limitations, large-scale GWAS represent a new tool for the discovery and the development of innovative drugs.