Methods for genetic epidemiology

Genome-wide association studies (GWAS) have proved successful in identifying new loci underlying susceptibility to a wide variety of traits including many cancers. It is becoming clear that in addition to relatively large genetic effects, such as pigmentation-related genetic effects on cancers such as melanoma, there exist a large number of smaller genetic effects which are more difficult to identify (“the polygenic tail” in the distribution of effects). Identifying and characterizing such smaller effects will require approaches which do more than simply focus on the very top hits from GWAS. Gene-based and pathway-based approaches are two approaches which complement single marker based GWAS analysis. We recently developed a versatile gene-based test for genome-wide association studies (GWAS)1. Our approach, called VEGAS (VErsatile Gene-based Association Study), is applicable to all GWAS designs, including family-based GWAS, meta-analyses of GWAS on the basis of summary data, and DNA-pooling-based GWAS, where existing approaches based on permutation are not possible, as well as singleton data, where they are. The test incorporates information from a full set of markers (or a defined subset) within a gene and accounts for linkage disequilibrium between markers by using simulations from the multivariate normal distribution. For an association study using singletons, our approach produces results equivalent to those obtained via permutation in a fraction of the computation time. Applying VEGAS to a DNA-pooling-based GWAS for melanoma (1200 cases with relevant controls), we previously showed that several pigmentation-related genes are clearly associated at the gene-based level. We have subsequently applied VEGAS to an extended melanoma data based on individually genotyped samples (2200 cases with relevant controls), obtaining stronger confirmation of the role of pigmentation related genes (MC1R, ASIP, OCA2, MTAP, TYR, SLC45A2) and suggestive evidence for a number of novel nonpigmentation related genes in melanoma. We also applied VEGAS to two further smaller (500 and 1000 cases, respectively) GWASs of melanoma, again highlighting some, but not all of the pigmentation genes listed above. Work is underway in applying VEGAS to several other cancer types including breast cancer, ovarian cancer, prostate cancer, pancreatic cancer and lung cancer. VEGAS has been extended to use gene-based results in a pathway analysis (VEGAS-pathway). VEGAS-pathway, like VEGAS, uses summary data and is not limited to situations where raw genotypes are available. Currently, this implementation combines gene-based test statistics across genes in a predefined pathway, taking into account situations where genes within the pathway are close to each other in the genome. This implementation has been evaluated using pathways defined by Gene Ontology and Ingenuity. Work is ongoing to extend the approach to further sets of predefined pathways. Applying VEGAS-pathway to the N=2200 melanoma data set using GO defined pathways, several pathways that made good biological sense (e.g., “melanin biosynthetic process,” “pigmentation”) were apparent. VEGAS-pathway results on the smaller (N=500, 1000) melanoma data sets did not clearly identify these GO pathways, indicating that large sample size remains crucial, even for situations such as pigmentation-related pathways in melanoma. Using Ingenuity-defined pathways, it was not possible to identify any statistically significant pathways that made good biological sense in any of the melanoma data sets – at least in this case, the more narrowly defined Ingenuity pathways offered less insight into the trait than the more broadly defined and numerous GO pathways. After removal of the known pigmentation genes, the most significant GO pathway in the N=2200 data set was the small pathway “regulation of growth rate”. This pathway showed nominal replication in a further set of N=1800 melanoma cases (with relevant controls). VEGAS-pathway results for several other cancer types (breast cancer, ovarian cancer, prostate cancer, pancreatic cancer and lung cancer) will be presented at the meeting. 1 Liu JZ, McRae AF, Nyholt DR, Medland SE, Wray NR, Brown KM, Hayward NK, Montgomery GW, Visscher PM, Martin NG, Macgregor S, A versatile gene-based test for genome-wide association studies. Am J Hum Genet 2010, 87(1):139-145. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr SY29-01. doi:10.1158/1538-7445.AM2011-SY29-01

Transethnic genome-wide meta-analysis for Alzheimer disease

Gene-based association tests aggregate genotypes across multiple variants for each gene, providing an interpretable gene-level analysis framework for genome-wide association studies (GWAS). Early gene-based test applications often focused on rare coding variants; a more recent wave of gene-based methods, e.g. TWAS, use eQTLs to interrogate regulatory associations. Regulatory variants are expected to be particularly valuable for gene-based analysis, since most GWAS associations to date are non-coding. However, identifying causal genes from regulatory associations remains challenging and contentious. Here, we present a statistical framework and computational tool to integrate heterogeneous annotations with GWAS summary statistics for gene-based analysis, applied with comprehensive coding and tissue-specific regulatory annotations. We compare power and accuracy identifying causal genes across single-annotation, omnibus, and annotation-agnostic gene-based tests in simulation studies and an analysis of 128 traits from the UK Biobank, and find that incorporating heterogeneous annotations in gene-based association analysis increases power and performance identifying causal genes.

IntroductionInsulin resistance (IR) is a key feature in the pathophysiology of metabolic disorders such as type 2 diabetes mellitus (T2DM) and obesity, both of which have a high prevalence in the Mexican population. Genetic predisposition plays a critical role in the development of IR, particularly through variants in carbohydrate metabolism genes. However, the specific contributions of these genetic factors in young Mexicans remain poorly characterized.ObjectivesThis study aimed to identify associations between single nucleotide polymorphisms (SNPs) in carbohydrate metabolism-related genes and insulin resistance markers in a young Mexican population. Additionally, we sought to identify novel genetic variants that may contribute to metabolic risk and explore sex-specific genetic effects.MethodsA genome-wide association study was conducted on 455 urban college students from Mexico. A total of 430 SNPs related to carbohydrate metabolism were genotyped. Clinical markers of IR, including fasting glucose, insulin, the homeostasis model assessment for insulin resistance (HOMA-IR), body mass index (BMI), body fat percentage (BF%), and waist circumference (WC), were evaluated. Associations between SNPs and metabolic traits were analyzed using logistic regression models adjusted for relevant covariates. Bonferroni correction was applied to account for multiple testing.ResultsSeveral SNPs were significantly associated with IR-related traits. SNPs in glucose metabolism genes, including GCK (rs1799884), SLC2 A9 (rs1122141), and IDH3B (rs6037255), were linked to fasting glucose levels, with IDH3B showing the strongest effect (OR 12.17, p = 0.0144). Variants in insulin-related genes, such as SLC2 A9, PFKP (rs3814591), and SLC45 A1 (rs12132135), were associated with elevated insulin and HOMA-IR. Adiposity-related SNPs, including those in PDK3 (rs7889665), PGK1 (rs2076630), and IDH2 (rs62019177), influenced body fat percentage and BMI. Notably, the novel variant rs77487659 (PGLS) was significantly associated with IR, particularly in women (OR 5.5, p = 0.00005), highlighting a potential sex-specific effect. The SLC45 A1 variant rs12132135 demonstrated significant associations with multiple metabolic traits, including BMI, WC, and insulin resistance.ConclusionThis study provides a comprehensive genetic analysis of IR-related traits in young Mexicans, highlighting both previously reported and novel associations. The findings suggest that genetic predisposition plays an important role in metabolic risk, with potential implications for early screening and personalized interventions. Sex-specific effects emphasize the need for tailored approaches in genetic risk assessment. Future research should focus on functional validation of these SNPs and their integration into precision medicine strategies for metabolic disease prevention.

Objectives: Both iron overload and iron deficiency cause oxidative stress and DNA damage, which increase carcinogenesis risk, especially in the gastrointestinal (GI) tract. In animal models, iron overload causes forestomach tumors, and in epidemiologic studies, high heme iron intake is associated with some upper GI cancers. Hemochromatosis, which results in excessive iron absorption in the GI tract, is also associated with GI cancer. The mechanism of increased cancer risk may involve the impairment of iron-dependent metabolic functions related to genome protection and maintenance. Genetic variants of genes with iron-dependent metabolic functions may confer differential GI cancer risk. Therefore, we comprehensively examined the genetic variants of these genes and risk of esophageal squamous cell carcinoma (ESCC) and gastric cardia adenocarcinoma (GCA) in a high-risk population in China. Methods: We genotyped 249 tag single nucleotide polymorphisms (SNPs) in 23 genes with known iron-dependent metabolic function in 1027 ESCC cases and 752 GCA cases plus 1452 controls from two epidemiologic studies in a high-risk region in China. For all genes as a single pathway and each gene individually, we calculated summary p-values using the adaptive rank truncated product (ARTP) method. For SNPs, logistic regression models were used to estimate odds ratios (OR) and 95% confidence intervals (CI). Results: In pathway-based tests, the iron pathway was not associated with either ESCC (p=0.320) or GCA (p=0.550). However, in gene-based tests, HMOX2 (heme oxygenase 2, p=0.018) and BMP2 (bone morphogenetic protein 2, p=0.032) were associated with ESCC risk, and IREB2 (iron-responsive element binding protein 2, p=0.033) was associated with GCA risk. In SNP-based tests, RS235756 (OR=0.78, CI=0.66-0.91) and RS910141 (OR=0.76, CI=0.63-0.93) tagged to BMP2 (linkage disequilibrium of pairwise r2=0.55) were associated with ESCC risk. RS6500609 (OR=1.30, CI=1.08-1.57) tagged to HMOX2 was also associated with ESCC risk. RS12910090 (OR=0.83, CI=0.72-0.94) and RS2568500 (OR=0.83, CI=0.73-0.95) tagged to IREB2 (r2=0.90) were associated with GCA risk. RS11630228 (OR=1.21, CI=1.06-1.38) also tagged to IREB2 (r2<0.38 with two previous SNPs) was associated with GCA risk. Conclusions: The potential link between iron status and differential UGI cancer risk may involve the impairment of iron-dependent metabolic functions, which leads to oxidative stress and DNA damage. The association of SNPs in HMOX2 and BMP2 with ESCC risk and IREB2 with GCA risk suggests that genetic variants of genes with iron-dependent metabolic functions may differentially confer risk of upper GI cancer. Further research is warranted to replicate these findings and elucidate the functional effects of the SNPs in the context of upper GI carcinogenesis to better understand the role of iron overload or iron deficiency in cancer development. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 2632. doi:1538-7445.AM2012-2632

BackgroundThe lack of integrated analysis of genome-wide association studies (GWAS) and 3D epigenomics restricts a deep understanding of the genetic mechanisms of meat-related traits. With the application of techniques as ChIP-seq and Hi-C, the annotations of cis-regulatory elements in the pig genome have been established, which offers a new opportunity to elucidate the genetic mechanisms and identify major genetic variants and candidate genes that are significantly associated with important economic traits. Among these traits, loin muscle depth (LMD) is an important one as it impacts the lean meat content. In this study, we integrated cis-regulatory elements and genome-wide association studies (GWAS) to identify candidate genes and genetic variants regulating LMD.ResultsFive single nucleotide polymorphisms (SNPs) located on porcine chromosome 17 were significantly associated with LMD in Yorkshire pigs. A 10 kb quantitative trait locus (QTL) was identified as a candidate functional genomic region through the integration of linkage disequilibrium and linkage analysis (LDLA) and high-throughput chromosome conformation capture (Hi-C) analysis. The BMP2 gene was identified as a candidate gene for LMD based on the integrated results of GWAS, Hi-C meta-analysis, and cis-regulatory element data. The identified QTL region was further verified through target region sequencing. Furthermore, through using dual-luciferase assays and electrophoretic mobility shift assays (EMSA), two SNPs, including SNP rs321846600, located in the enhancer region, and SNP rs1111440035, located in the promoter region, were identified as candidate SNPs that may be functionally related to the LMD.ConclusionsBased on the results of GWAS, Hi-C, and cis-regulatory elements, the BMP2 gene was identified as an important candidate gene regulating variation in LMD. The SNPs rs321846600 and rs1111440035 were identified as candidate SNPs that are functionally related to the LMD of Yorkshire pigs. Our results shed light on the advantages of integrating GWAS with 3D epigenomics in identifying candidate genes for quantitative traits. This study is a pioneering work for the identification of candidate genes and related genetic variants regulating one key production trait (LMD) in pigs by integrating genome-wide association studies and 3D epigenomics.

Genome-wide association study and genomic prediction of Fusarium ear rot resistance in tropical maize germplasm

Schizophrenia is a chronic and debilitating psychiatric condition affecting slightly more than 1% of the population worldwide and it is a multifactorial disorder with a high degree of heritability (80%) based on family and twin studies. Increasing lines of evidence suggest intermediate phenotypes/endophenotypes are more associated with causes of the disease and are less genetically complex than the broader disease spectrum. Negative symptoms in schizophrenia are attractive intermediate phenotypes based on their clinical and treatment response features. Therefore, our objective was to identify genetic variants underlying the negative symptoms of schizophrenia by analyzing two genome-wide association (GWA) data sets consisting of a total of 1,774 European-American patients and 2,726 controls. Logistic regression analysis of negative symptoms as a binary trait (adjusted for age and sex) was performed using PLINK. For meta-analysis of two datasets, the fixed-effect model in PLINK was applied. Through meta-analysis we identified 25 single nucleotide polymorphisms (SNPs) associated with negative symptoms with p<5×10−5. Especially we detected five SNPs in the first two genes/loci strongly associated with negative symptoms of schizophrenia (Pmeta-analysis<6.22×10−6), which included three SNPs in the BCL9 gene: rs583583 showed the strongest association at a Pmeta-analysis of 6.00×10−7 and two SNPs in the C9orf5 (the top SNP is rs643410 with a p = 1.29 ×10−6). Through meta-analysis, we identified several additional negative symptoms associated genes (ST3GAL1, RNF144, CTNNA3 and ZNF385D). This is the first report of the common variants influencing negative symptoms of schizophrenia. These results provide direct evidence of using of negative symptoms as an intermediate phenotype to dissect the complex genetics of schizophrenia. However, additional studies are warranted to examine the underlying mechanisms of these disease-associated SNPs in these genes.

Alterations of activity and expression in tricarboxylic acid (TCA) cycle key enzymes have been indicated in several malignancies, including hepatocellular carcinoma (HCC). They play an important role in the progression of cancer. However, the impact of single nucleotide polymorphisms (SNPs) in genes encoding these key enzymes on the recurrence of HCC has not been investigated. In this study, we genotyped 17 SNPs in genes encoding TCA cycle key enzymes and analyzed their association with recurrence-free survival (RFS) in a cohort of 492 Chinese HCC patients by Cox proportional hazard model and survival tree analysis. We identified 7 SNPs in SDHC, SDHD, FH, and IDH2 genes to be significantly associated with the RFS of HCC patients. Moreover, all these SNPs were associated with the early recurrence (within 2 years after surgery) risk of diseases. Cumulative effect analysis showed that these SNPs exhibited a dose-dependent effect on the overall and early recurrence. Further stratified analysis suggested that number of risk genotypes modified the protective effect on HCC recurrence conferred by transcatheter arterial chemoembolization treatment. Finally, the survival tree analysis revealed that SNP rs10789859 in SDHD gene was the primary factor contributing to HCC recurrence in our population. To the best of our knowledge, we for the first time observed the association between SNPs in genes encoding TCA cycle key enzymes and HCC recurrence risk. Further observational and functional studies are needed to validate our findings and generalize its clinical usage.

Introduction: In this study, we evaluated the genetic predisposition to AKI using genome-wide association study (GWAS) in patients undergoing elective off-pump-CABG (OP-CABG). Methods: Patients of South Asian ancestry were categorized into four cohorts as follows: (A) patients with normal renal function undergoing OP-CABG (eGFR&gt;60ml/min/m2) (n=754; AKI - 81, Non-AKI-673); (B) patients with renal dysfunction (eGFR&lt;60ml/min/m2; not needing dialysis) undergoing OP-CABG (n=263; AKI-169, non-AKI-89); (C) patients with renal dysfunction with no symptomatic/known heart disease (n=92); (D) age-matched healthy control population (with normal kidney function, without symptomatic/known heart disease) (n=826). Renal function was considered to be normal, if the serum creatinine was less than 1.3mg% and/or estimated glomerular filtration rate was greater than 60ml/min/1.7 m 2 . Genotyping was performed using Infinium Global Screening Array-24 v3.0 BeadChip from Illumina, which included 6,54,027 markers. GWAS analysis was performed using age, genders and top10 principal components as covariates using logistic regression using PLINK 2.0. Postoperative AKI was defined based on Kidney Disease Improving Global Outcomes guidelines. Results: There were 754 participants with good renal function undergoing OP-CABG; among them, 81 subjects developed AKI postoperatively. The GWAS analysis revealed 607 markers with a suggestive significance p-value threshold of 1e-4. A modest signal was observed on chromosome 11 region, covering genes such as COX8A, NAA40, RCOR. The gene-based test in FUMA annotation revealed genes, MARCOD2, CNTN4, MAGI2 etc. as top hits. However, none of the markers remained significant after multiple testing corrections using Bonferroni and Benjamini-Hochberg. Conclusions: The GWAS revealed an association signal on chromosome 11, which is suggestive and needs to be pursued further in a larger sample set.

We examined the associations of single nucleotide polymorphisms (SNPs) in three candidate hypertension genes, alpha-adducin (ADD1/G460W), beta2-adrenergic receptor (ADRB2/Arg16Gly and Gln27Glu) and G-protein beta3 subunit (GNB3/C825T), with retinal arteriolar calibre (an intermediate marker of chronic hypertension) and venular calibre. Data in 1842 participants (1554 whites and 288 African Americans) aged 69-96 years from the Cardiovascular Health Study with genotype and retinal vascular calibre data were included. A computer-assisted method was used to measure retinal vascular calibre. We analysed four SNPs and multilocus interaction for three genes. All SNPs were in Hardy-Weinberg equilibrium in whites and African Americans. The study had sufficient power to detect 0.5% of the total variance of retinal vascular calibre contributed by each SNP in the total population, except for the GNB3 gene variant. No significant associations between these SNPs in the genes studied and mean retinal arteriolar and venular calibre were found in single-gene or multilocus analysis (for example, age-, gender-, race-adjusted mean retinal arteriolar calibre was similar between participants who were ADD1/460W homozygotes and ADD1/G allele carriers, 166.2 vs 167.7 microm). In conclusion, this study found no evidence of an association of SNPs in candidate hypertension genes studied here with retinal vascular calibre.

In chapter one, we proposed a novel multiple phenotype association test methods to test the association between multiple phenotypes and a SNP. Genome-wide association studies (GWAS) have identified many strongly associated genetic variants with phenotypes and have greatly enhanced our understanding of the genetic architecture of complex phenotypes and diseases. In this study, we proposed a new method to construct a Phenotype-Phenotype Network using the sparse Gaussian Graphical Model (sGGM) based on GWAS summary statistics. This approach will isolate the direct relationship between phenotypes, making it easier to identify clusters of phenotypes conditional on other phenotypes that reflect the more meaningful biological or functional connections. Then we applied the community detection method to partition phenotypes into disjoint modules based on the partial correlation matrix of phenotypes and applied multiple phenotype test methods on each of the clusters and combined p-values to get the final P-value. The comprehensive simulation study showed that our method can control the type I error rate effectively and has the highest power compared to the other methods we compared. Application of this method to the GWAS summary statistics of 92 phenotypes from UK-Biobank data has identified higher number of significant SNPs than other methods. Downstream analysis of the significant SNPs shows the functional and biological importance. The complex traits are often the result of the combined effect of multiple genetic variants, each with a small individual effect and these small effects may be missed due to stringent p-value thresholds in GWAS studies. Gene-based association tests combine the effects of multiple variants within a gene and provide more interpretable associations and enable researchers to detect associations that might be missed in single-SNP analyses. Gene-based tests increase statistical power by considering the cumulative effect of several variants in a gene, reducing the multiple testing burden and prioritizing variants based on their likelihood of contributing to the phenotype. The power of the gene-based test depended on the genetic architecture of the trait and the weight of SNPs in a gene. The genetic architecture of the complex trait is not known in advance and the power of the gene-based association tests decreases if the weight is mis-specified. The equal weighting assumption implies that all SNPs within the gene have an equal effect on the trait. On the other hand, minor allele frequency (MAF)-based weights, such as inverse or beta distribution of MAF, upweight rare variants under the assumption that they

Repurposing existing medicines to target disease-associated genes represents a promising strategy for developing effective treatments for complex diseases. However, progress has been hindered by a lack of viable candidate drug targets identified through genome-wide association studies. Gene-based association tests provide a more powerful alternative to traditional SNP-based methods, yet current approaches often fail to leverage shared heritability across populations and to effectively integrate functional genomic data. To address these challenges, we develop GenT and its various extensions, comprising a framework of gene-based tests utilizing summary-level data from genome-wide association studies. Using GenT, we identify 16, 15, 35, and 83 candidate genes linked to Alzheimer's disease, amyotrophic lateral sclerosis, major depression, and schizophrenia, respectively, not detected by Genome-Wide Association Studies (GWAS). Additionally, we use our multi-ancestry gene-based test (MuGenT) to identify 28 candidate genes associated with type 2 diabetes. By integrating brain expression and protein quantitative trait loci into our analysis, we identify 43 candidate genes associated with Alzheimer's disease that have supporting xQTL evidence. We also perform experimental assays to demonstrate that the NTRK1 inhibitor GW441756 significantly reduces tau hyper-phosphorylation (including p-tau181 and p-tau217) in Alzheimer's disease patient-derived iPSC neurons, providing mechanistic support for our predictions.

Genome-wide association studies (GWAS) have successfully identified many common genetic variants associated with complex diseases over the past decade. The 'gold standard' method for validating the top single nucleotide polymorphisms (SNPs) identified in GWAS is to independently replicate the findings in similar or diverse large-scale external cohorts. However, for rare diseases, it can be difficult to find an external validation cohort within a reasonable timeframe. In such situations, resampling methods, such as the two-step iterative resampling (TSIR) approach have been used to identify SNPs associated with the outcome of interest. However, the TSIR approach involves choosing several parameters in each step, which can influence the performance of the approach. In this paper, we undertook extensive simulation studies to assess the effect of choice of different parameters on the type I error and power for both binary and continuous phenotypes and also compared the TSIR approach with the traditional one-stage (OS) and two-stage (TS) GWAS analysis. We illustrate the usefulness of the TSIR approach by applying it to a GWAS of childhood cancer survivors. Our results indicate that the TSIR approach with an at least 70:30 split and a cutoff of discovering and replicating SNPs at least 20 times in 100 replications provides conservative type I error control and has near 'optimal' power for internally validated SNPs. Its performance is comparable with the TS GWAS for which an external validation cohort is available with only slight reduction in power in some situations. It has almost the same power as OS GWAS with conservative type I error which leads to fewer false positive findings. TSIR is a powerful and efficient method for identifying and internally validating SNPs for GWAS when independent cohorts for external validation may not be available.

The β2-adrenoceptor gene and hypertension: is it the promoter or the coding region or neither?