Comparison of population- and family-based methods for genetic association analysis in the presence of interacting loci

Selecting cases and controls for DNA sequencing studies using family histories of disease

Recent improvements in sequencing technology have enabled the investigation of so-called missing heritability, and a large number of affected subjects have been sequenced in order to detect significant associations between human diseases and rare variants. However, the cost of genome sequencing is still high, and a statistically powerful strategy for selecting informative subjects would be useful. Therefore, in this report, we propose a new statistical method for selecting cases and controls for sequencing studies based on family history. We assume that disease status is determined by unobserved liability scores. Our method consists of two steps: first, the conditional means of liability are estimated with the liability threshold model given the individual's disease status and those of their relatives. Second, the informative subjects are selected with the estimated conditional means. Our simulation studies showed that statistical power is substantially affected by the subject selection strategy chosen, and power is maximized when affected (unaffected) subjects with high (low) risks are selected as cases (controls). The proposed method was successfully applied to genome-wide association studies for type 2 diabetes, and our analysis results reveal the practical value of the proposed methods. Copyright © 2017 John Wiley & Sons, Ltd.

Confirmation of novel type 1 diabetes risk loci in families

AIMS/HYPOTHESIS

Over 50 regions of the genome have been associated with type 1 diabetes risk, mainly using large case/control collections. In a recent genome-wide association (GWA) study, 18 novel susceptibility loci were identified and replicated, including replication evidence from 2,319 families. Here, we, the Type 1 Diabetes Genetics Consortium (T1DGC), aimed to exclude the possibility that any of the 18 loci were false-positives due to population stratification by significantly increasing the statistical power of our family study.

METHODS

We genotyped the most disease-predicting single-nucleotide polymorphisms at the 18 susceptibility loci in 3,108 families and used existing genotype data for 2,319 families from the original study, providing 7,013 parent-child trios for analysis. We tested for association using the transmission disequilibrium test.

RESULTS

Seventeen of the 18 susceptibility loci reached nominal levels of significance (p < 0.05) in the expanded family collection, with 14q24.1 just falling short (p = 0.055). When we allowed for multiple testing, ten of the 17 nominally significant loci reached the required level of significance (p < 2.8 × 10(-3)). All susceptibility loci had consistent direction of effects with the original study.

CONCLUSIONS/INTERPRETATION

The results for the novel GWA study-identified loci are genuine and not due to population stratification. The next step, namely correlation of the most disease-associated genotypes with phenotypes, such as RNA and protein expression analyses for the candidate genes within or near each of the susceptibility regions, can now proceed.

Whole-genome linkage and association scan in primary, nonsyndromic vesicoureteric reflux

Primary vesicoureteric reflux accounts for approximately 10% of kidney failure requiring dialysis or transplantation, and sibling studies suggest a large genetic component. Here, we report a whole-genome linkage and association scan in primary, nonsyndromic vesicoureteric reflux and reflux nephropathy. We used linkage and family-based association approaches to analyze 320 white families (661 affected individuals, generally from families with two affected siblings) from two populations (United Kingdom and Slovenian). We found modest evidence of linkage but no clear overlap with previous studies. We tested for but did not detect association with six candidate genes (AGTR2, HNF1B, PAX2, RET, ROBO2, and UPK3A). Family-based analysis detected associations with one single-nucleotide polymorphism (SNP) in the UK families, with three SNPs in the Slovenian families, and with three SNPs in the combined families. A case-control analysis detected associations with three additional SNPs. The results of this study, which is the largest to date investigating the genetics of reflux, suggest that major loci may not exist for this common renal tract malformation within European populations.

Analysis of 19 genes for association with type I diabetes in the Type I Diabetes Genetics Consortium families

In recent years the pace of discovery of genetic associations with type I diabetes (T1D) has accelerated, with the total number of confirmed loci, including the major histocompatibility complex (MHC) region, reaching 43. However, much of the deciphering of the associations at these, and the established T1D loci, has yet to be performed in sufficient numbers of samples or with sufficient markers. Here, 257 single-nucleotide polymorphisms (SNPs) have been genotyped in 19 candidate genes (INS, PTPN22, IL2RA, CTLA4, IFIH1, SUMO4, VDR, PAX4, OAS1, IRS1, IL4, IL4R, IL13, IL12B, CEACAM21, CAPSL, Q7Z4c4(5Q), FOXP3, EFHB) in 2300 affected sib-pair families and tested for association with T1D as part of the Type I Diabetes Genetics Consortium's candidate gene study. The study had approximately 80% power at alpha=0.002 and a minor allele frequency of 0.2 to detect an effect with a relative risk (RR) of 1.20, which drops to just 40% power for a RR of 1.15. At the INS gene, rs689 (-23 HphI) was the most associated SNP (P=3.8 x 10(-31)), with the estimated RR=0.57 (95% confidence interval, 0.52-0.63). In addition, rs689 was associated with age-at-diagnosis of T1D (P=0.001), with homozygosity for the T1D protective T allele, delaying the onset of T1D by approximately 2 years in these families. At PTPN22, rs2476601 (R620W), in agreement with previous reports, was the most significantly associated SNP (P=6.9 x 10(-17)), with RR=1.55 (1.40-1.72). Evidence for association with T1D was observed for the IFIH1 SNP, rs1990760 (P=7.0 x 10(-4)), with RR=0.88 (0.82-0.95) and the CTLA4 SNP rs1427676 (P=0.0005), with RR=1.14 (1.06-1.23). In contrast, no convincing evidence of association was obtained for SUMO4, VDR, PAX4, OAS1, IRS1, IL4, IL4R, IL13, IL12B, CEACAM21 or CAPSL gene regions (http://www.T1DBase.org).

Interaction of gender, hypertension, and the angiotensinogen gene haplotypes on the risk of coronary artery disease in a large angiographic cohort

There is increasing evidence suggesting the importance of evaluating gene-environment interactions in the genetic study of coronary artery disease (CAD). We investigated the association of multiple single nucleotide polymorphisms in the angiotensinogen (AGT) gene with CAD, considering the interaction between the genetic and non-genetic factors, using a larger and ethnically homogeneous angiographic cohort. A total of 1254 consecutive patients who underwent cardiac catheterization (735 with CAD and 519 without) were recruited. T174M (rs4762), M235T (rs699), G-6A, A-20C, G-152A, and G-217A polymorphisms of the AGT gene were genotyped. We used a regression approach based on a generalized linear model to evaluate haplotype effects defined by the multilocus data and detection of gene-environment interaction by incorporating interaction terms in the model. We found significant differences in global AGT gene haplotype profile between patients with and without CAD (the global score statistic=25.411, P=0.008). Significant interactions between AGT gene haplotypes, gender and hypertension were detected. We also used haplotype counting to directly estimate the odds ratio of each AGT gene haplotype, and found that the effects of haplotypes were markedly different in subgroups defined by gender and hypertension, providing strong evidence of gene-environment interaction. Female gender synergistically enhances (or male gender reverses) the effects of AGT gene haplotypes on the risk of CAD in the presence of hypertension. In conclusion, the effect of AGT gene haplotypes on the risk of CAD was significantly increased in women with hypertension, which highlights the importance of evaluating gene-environment interactions in the genetic study of CAD.

PTPN22 Trp620 explains the association of chromosome 1p13 with type 1 diabetes and shows a statistical interaction with HLA class II genotypes

OBJECTIVE

The disease association of the common 1858C>T Arg620Trp (rs2476601) nonsynonymous single nucleotide polymorphism (SNP) of protein tyrosine phosphatase; nonreceptor type 22 (PTPN22) on chromosome 1p13 has been confirmed in type 1 diabetes and also in other autoimmune diseases, including rheumatoid arthritis and Graves' disease. Some studies have reported additional associated SNPs independent of rs2476601/Trp(620), suggesting that it may not be the sole causal variant in the region and that the relative risk of rs2476601/Trp(620) is greater in lower risk by HLA class II genotypes than in the highest risk class II risk category.

RESEARCH DESIGN AND METHODS

We resequenced PTPN22 and used these and other data to provide >150 SNPs to evaluate the association of the PTPN22 gene and its flanking chromosome region with type 1 diabetes in a minimum of 2,000 case subjects and 2,400 control subjects.

RESULTS

Due to linkage disequilibrium, we were unable to distinguish between rs2476601/Trp(620) (P = 2.11 x10(-87)) and rs6679677 (P = 3.21 x10(-87)), an intergenic SNP between the genes putative homeodomain transcription factor 1 and round spermatid basic protein 1. None of the previously reported disease-associated SNPs proved to be independent of rs2476601/Trp(620). We did not detect any interaction with age at diagnosis or sex. However, we found that rs2476601/Trp(620) has a higher relative risk in type 1 diabetic case subjects carrying lower risk HLA class II genotypes than in those carrying higher risk ones (P = 1.36 x 10(-4) in a test of interaction).

CONCLUSIONS

In our datasets, there was no evidence for allelic heterogeneity at the PTPN22 locus in type 1 diabetes, indicating that the SNP rs2476601/Trp(620) remains the best candidate in this chromosome region in European populations. The heterogeneity of rs2476601/Trp(620) disease risk by HLA class II genotype is consistent with previous studies, and the joint effect of the two loci is still greater in the high-risk group.

Genome-wide association studies for complex traits: consensus, uncertainty and challenges

The past year has witnessed substantial advances in understanding the genetic basis of many common phenotypes of biomedical importance. These advances have been the result of systematic, well-powered, genome-wide surveys exploring the relationships between common sequence variation and disease predisposition. This approach has revealed over 50 disease-susceptibility loci and has provided insights into the allelic architecture of multifactorial traits. At the same time, much has been learned about the successful prosecution of association studies on such a scale. This Review highlights the knowledge gained, defines areas of emerging consensus, and describes the challenges that remain as researchers seek to obtain more complete descriptions of the susceptibility architecture of biomedical traits of interest and to translate the information gathered into improvements in clinical management.

Power of genome-wide association studies in the presence of interacting loci

Though multiple interacting loci are likely involved in the etiology of complex diseases, early genome-wide association studies (GWAS) have depended on the detection of the marginal effects of each locus. Here, we evaluate the power of GWAS in the presence of two linked and potentially associated causal loci for several models of interaction between them and find that interacting loci may give rise to marginal relative risks that are not generally considered in a one-locus model. To derive power under realistic situations, we use empirical data generated by the HapMap ENCODE project for both allele frequencies and linkage disequilibrium (LD) structure. The power is also evaluated in situations where the causal single nucleotide polymorphisms (SNPs) may not be genotyped, but rather detected by proxy using a SNP in LD. A common simplification for such power computations assumes that the sample size necessary to detect the effect at the tSNP is the sample size necessary to detect the causal locus directly divided by the LD measure r(2) between the two. This assumption, which we call the "proportionality assumption", is a simplification of the many factors that contribute to the strength of association at a marker, and has recently been criticized as unreasonable (Terwilliger and Hiekkalinna [2006] Eur J Hum Genet 14(4):426-437), in particular in the presence of interacting and associated loci. We find that this assumption does not introduce much error in single locus models of disease, but may do so in so in certain two-locus models.

Modeling gene-gene interactions using graphical chain models

OBJECTIVE

To investigate whether graphical chain models are suitable to detect gene-gene interaction under different biological models.

METHODS

We conducted a simulation study comparing graphical chain models with logistic regression models regarding their ability to detect underlying biological interaction models. For both methods, we attempted to capture simulation data following 12 different biological models. We used 10 statistical models for both methods. Of the 12 different biological models, four contained no interaction effects, two were multiplicative, and six were epistasis models. For each situation, the choice for a statistical model was based on global model fit as judged by two different information criteria, the BIC and the AIC.

RESULTS

Both methods failed in most of the scenarios to capture the gene-gene interaction present in the simulation data. Only in very specific cases, when disease risk was high and both genes had a dominant effect, present gene-gene interaction was detected.

CONCLUSIONS

Graphical chain models are, similar to logistic regression models, not able to capture gene-gene interactions for arbitrary biological models underlying the data.

Family based studies and genetic epidemiology: theory and practice

Family based studies have underpinned many successes in uncovering the causes of monogenic and oligogenic diseases. Now research is focussing on the identification and characterisation of genes underlying common diseases and it is widely accepted that these studies will require large population based samples. Population based family study designs have the potential to facilitate the analysis of the effects of both genes and environment. These types of studies integrate the population based approaches of classic epidemiology and the methods enabling the analysis of correlations between relatives sharing both genes and environment. The extent to which such studies are feasible will depend upon population- and disease-specific factors. To review this topic, a symposium was held to present and discuss the costs, requirements and advantages of population based family study designs. This article summarises the features of the meeting held at The University of Sheffield, August 2006.

A type 1 diabetes subgroup with a female bias is characterised by failure in tolerance to thyroid peroxidase at an early age and a strong association with the cytotoxic T-lymphocyte-associated antigen-4 gene

AIMS/HYPOTHESIS

HLA haplotypes DRB1*03_DQB1*02 and DRB1*04_DQB1*0302, and allelic variation of the T cell regulatory gene cytotoxic T-lymphocyte-associated antigen-4 (CTLA4) and of the T cell activation gene protein tyrosine phosphatase, non-receptor type 22 (lymphoid) (PTPN22) have been associated with type 1 diabetes and autoimmune thyroid disease. Using thyroid peroxidase autoantibodies (TPOAbs) as an indicator of thyroid autoimmunity, we assessed whether the association of these loci is different in type 1 diabetes patients with TPOAbs than in those without.

MATERIALS AND METHODS

TPOAbs were measured in 4,364 type 1 diabetic patients from across Great Britain, 67% of whom were aged under 18 years. These patients and 6,866 geographically matched control subjects were genotyped at CTLA4, PTPN22, HLA-DRB1 and HLA-DQB1.

RESULTS

TPOAbs were detected in 462 (10.6%) of the type 1 diabetic patients. These patients had a stronger association with CTLA4 (odds ratio [OR] = 1.49 for the G allele of the single nucleotide polymorphism rs3087243; 95% CI = 1.29-1.72) than did the TPOAbs-negative patients (p = 0.0004; OR = 1.16; 95% CI = 1.10-1.24) or type 1 diabetes patients overall (OR = 1.20; 95% CI = 1.13-1.27). The ratio of women:men was higher (1.94:1) in this subgroup than in type 1 diabetes patients without TPOAbs (0.94:1; p = 1.86 x 10(-15)). TPOAbs status did not correlate with age at diagnosis of type 1 diabetes or with PTPN22 (Arg620Trp; rs2476601).

CONCLUSIONS/INTERPRETATION

Our results identify a subgroup of type 1 diabetic patients that is sensitive to allelic variation of the negative regulatory molecule CTLA-4 and indicate that TPOAbs testing could be used to subclassify type 1 diabetes patients for inclusion in genetic, biological or clinical studies.

Ascertainment through family history of disease often decreases the power of family-based association studies

Selection of cases with additional affected relatives has been shown to increase the power of the case-control association design. We investigated whether this strategy can also improve the power of family-based association studies that use the transmission disequilibrium test (TDT), while accounting for the effects of residual polygenic and environmental factors on disease liability. Ascertainment of parent-offspring trios conditional on the proband having affected first-degree relatives almost always reduced the power of the TDT. For many disease models, this reduction was quite considerable. In contrast, for the same sample size, designs that analyzed more than one affected offspring per family often improved power when compared to the standard parent-offspring trio design. Together, our results suggest that (1) residual polygenic and environmental influences should be considered when estimating the power of the TDT for studies that ascertain families with multiple affected relatives; (2) if trios are selected conditional on having additional affected offspring, then it is important to genotype and include in the analysis the additional siblings; (3) the ascertainment strategy should be considered when interpreting results from TDT analyses. Our analytic approach to estimate the asymptotic power of the TDT is implemented online at http://pngu.mgh.harvard.edu/ ~purcell/gpc/.

CT60 genotype does not affect CTLA-4 isoform expression despite association to T1D and AITD in northern Sweden

BACKGROUND

Polymorphisms in and around the CTLA-4 gene have previously been associated to T1D and AITD in several populations. One such single nucleotide polymorphism (SNP), CT60, has been reported to affect the expression level ratio of the soluble (sCTLA-4) to full length CTLA-4 (flCTLA-4) isoforms. The aims of our study were to replicate the association previously published by Ueda et al. of polymorphisms in the CTLA-4 region to T1D and AITD and to determine whether the CT60 polymorphism affects the expression level ratio of sCTLA-4/flCTLA-4 in our population.

METHODS

Three SNPs were genotyped in 253 cases (104 AITD cases and 149 T1D cases) and 865 ethnically matched controls. Blood from 23 healthy individuals was used to quantify mRNA expression of CTLA-4 isoforms in CD4+ cells using real-time PCR. Serum from 102 cases and 59 healthy individuals was used to determine the level of sCTLA-4 protein.

RESULTS

Here we show association of the MH30, CT60 and JO31 polymorphisms to T1D and AITD in northern Sweden. We also observed a higher frequency of the CT60 disease susceptible allele in our controls compared to the British, Italian and Dutch populations, which might contribute to the high frequency of T1D in Sweden. In contrast to previously published findings, however, we were unable to find differences in the sCTLA-4/flCTLA-4 expression ratio based on the CT60 genotype in 23 healthy volunteers, also from northern Sweden. Analysis of sCTLA-4 protein levels in serum showed no correlation between sCTLA-4 protein levels and disease status or CT60 genotype.

CONCLUSION

Association was found between T1D/AITD and all three polymorphisms investigated. However, in contrast to previous investigations, sCTLA-4 RNA and protein expression levels did not differ based on CT60 genotype. Our results do not rule out the CT60 SNP as an important polymorphism in the development of T1D or AITD, but suggest that further investigations are necessary to elucidate the effect of the CTLA-4 region on the development of T1D and AITD.

Interaction analysis of the CBLB and CTLA4 genes in type 1 diabetes

Gene-gene interaction analyses have been suggested as a potential strategy to help identify common disease susceptibility genes. Recently, evidence of a statistical interaction between polymorphisms in two negative immunoregulatory genes, CBLB and CTLA4, has been reported in type 1 diabetes (T1D). This study, in 480 Danish families, reported an association between T1D and a synonymous coding SNP in exon 12 of the CBLB gene (rs3772534 G>A; minor allele frequency, MAF=0.24; derived relative risk, RR for G allele=1.78; P=0.046). Furthermore, evidence of a statistical interaction with the known T1D susceptibility-associated CTLA4 polymorphism rs3087243 (laboratory name CT60, G>A) was reported (P<0.0001), such that the CBLB SNP rs3772534 G allele was overtransmitted to offspring with the CTLA4 rs3087243 G/G genotype. We have, therefore, attempted to obtain additional support for this finding in both large family and case-control collections. In a primary analysis, no evidence for an association of the CBLB SNP rs3772534 with disease was found in either sample set (2162 parent-child trios, P=0.33; 3453 cases and 3655 controls, P=0.69). In the case-only statistical interaction analysis between rs3772534 and rs3087243, there was also no support for an effect (1994 T1D affected offspring, and 3215 cases, P=0.92). These data highlight the need for large, well-characterized populations, offering the possibility of obtaining additional support for initial observations owing to the low prior probability of identifying reproducible evidence of gene-gene interactions in the analysis of common disease-associated variants in human populations.

CFH, ELOVL4, PLEKHA1 and LOC387715 genes and susceptibility to age-related maculopathy: AREDS and CHS cohorts and meta-analyses

Age-related maculopathy (ARM) is an important cause of visual impairment in the elderly population. It is of crucial importance to identify genetic factors and their interactions with environmental exposures for this disorder. This study was aimed at investigating the CFH, ELOVL4, PLEKHA1 and LOC387715 genes in independent cohorts collected using different ascertainment schemes. The study used a case-control design with subjects originally recruited through the Cardiovascular Health Study (CHS) and the Age-Related Eye Disease Study (AREDS). CFH was significantly associated with ARM in both cohorts (P</=0.00001). A meta-analysis confirmed that the risk allele in the heterozygous or homozygous state (OR, 2.4 and 6.2; 95% CI, 2.2-2.7 and 5.4-7.2, respectively) confers susceptibility. LOC387715 was also significantly associated with ARM in both cohorts (P</=0.00001) and a meta-analysis confirmed that the risk allele in the heterozygous and homozygous state (OR, 2.5 and 7.3; 95% CI, 2.2-2.9 and 5.7-9.4, respectively) confers susceptibility. Both CFH and LOC387715 showed an allele-dose effect on the ARM risk, individuals homozygous at either locus were at more than two-fold risk compared to those heterozygous. PLEKHA1, which is closely linked to LOC387715, was significantly associated with ARM status in the AREDS cohort, but not the CHS cohort and ELOVL4 was not significantly associated with ARM in either cohort. Joint action of CFH and LOC387715 was best described by independent multiplicative effect without significant interaction in both cohorts. Interaction of both genes with cigarette smoking was insignificant in both cohorts. This study provides additional support for the CFH and LOC387715 genes in ARM susceptibility via the evaluation of cohorts that had different ascertainment schemes regarding ARM status and through the meta-analyses.

Efficient study designs for test of genetic association using sibship data and unrelated cases and controls

Linkage mapping of complex diseases is often followed by association studies between phenotypes and marker genotypes through use of case-control or family-based designs. Given fixed genotyping resources, it is important to know which study designs are the most efficient. To address this problem, we extended the likelihood-based method of Li et al., which assesses whether there is linkage disequilibrium between a disease locus and a SNP, to accommodate sibships of arbitrary size and disease-phenotype configuration. A key advantage of our method is the ability to combine data from different family structures. We consider scenarios for which genotypes are available for unrelated cases, affected sib pairs (ASPs), or only one sibling per ASP. We construct designs that use cases only and others that use unaffected siblings or unrelated unaffected individuals as controls. Different combinations of cases and controls result in seven study designs. We compare the efficiency of these designs when the number of individuals to be genotyped is fixed. Our results suggest that (1) when the disease is influenced by a single gene, the one sibling per ASP-control design is the most efficient, followed by the ASP-control design, and familial cases contribute more association information than singleton cases; (2) when the disease is influenced by multiple genes, familial cases provide more association information than singleton cases, unless the effect of the locus being tested is much smaller than at least one other untested disease locus; and (3) the case-control design can be useful for detecting genes with small effect in the presence of genes with much larger effect. Our findings will be helpful for researchers designing and analyzing complex disease-association studies and will facilitate genotyping resource allocation.

Genetic susceptibility to type 1 diabetes

The recent discovery of PTPN22 as a novel susceptibility gene in human type 1 diabetes and continued progress in defining genes in animal models of the disease mark a fruitful period in the field of type 1 diabetes genetics. In addition, the similarities of the genetic and functional aspects across species have been substantiated. Future genome-wide association studies will reveal more loci, each providing a piece to the genetic puzzle of autoimmune disease.