Variants In Linkage Disequilibrium Research Articles

Technological advancements in functional interpretation of genome-wide association studies (GWAS) findings: bridging the gap to clinical translation.

Genome-wide association studies (GWAS) significantly advanced our understanding of the genetic underpinnings of diseases. However, challenges persist, particularly in interpreting non-coding variants in linkage disequilibrium that affect genes in disease-relevant cells. Addressing key obstacles-identifying causal variants, uncovering target genes, and understanding their network impact-is crucial. This graphical review navigates advanced techniques to fully leverage GWAS for future therapeutic breakthroughs.

Abstract 4887: Unveiling the functional impact of glioma non-coding germline variants using large language models

Abstract Efforts to identify the risk for developing gliomas through Genome-Wide Association Studies (GWAS) face relevant challenges, such as accurately pinpointing the genomic positions of germline variants, primarily found in non-coding regions, with unclear functional implications in glioma development. In this context, the GET method, a large language model designed to unveil regulatory grammars across 213 human cell types, emerges as a valuable tool for elucidating the mechanisms of non-coding mutations. Our aims include: I) Validate the efficacy of the GET method using a well-documented case and prioritize relevant variants for analysis; II) Refine the genomic positions of these variants and exploring their linkage disequilibrium (LD); III) Identify potential motifs and genes that might be influenced by these high-impact variants. After examining published GWAS, 39 non-coding SNPs in gliomas were found. Adding 316 variants in high LD (R2 > 0.8) expanded the dataset to 355 variants for analysis. Using the GET method, we calculated impact scores (ISs) for each variant, considering changes in motif binding score (ΔBM) and motif gradient (GM) across human astrocytes. We focused on rare variants (risk allele frequency <0.25) with the higher odds ratios (OR). The table presents GET results for a well-known case-control, showing its role in increasing glioma risk through OCT4 motif disruption and enhanced MYC expression, as previously was confirmed experimentally. Moreover, it details results concerning the 5 rare germline variants exhibiting the highest ISs, including precise genomic positions and the most affected motifs and genes that could potentially be impacted by each variant. Using a large language model, this study overcomes the challenge of mere variant identification, aiming to establish a framework for uncovering the underlying mechanism of risk variants. While further validation is essential, this is a promising approach to explore the impact of germline mutations. Variant ID Genomic Position Altered Motif Affected Gene Ref allele Alt allele Case-control rs55705857 Chr8:129633445 OCT4+SOX2:POU MYC A G H-I variant #1 rs78378222 Chr17:7668433 HD/18 PER1 T G H-I variant #2 rs518394 Chr9:22019673 KLF/SP/2 CDKN2A G C H-I variant #3 rs79495512 Chr15:76199995 KLF/SP/2 LINGO1 T C H-I variant #4 rs78185702 Chr15:76317893 SMAD LINGO1 C T H-I variant #5 rs2538067 Chr7:54907724 CTCF EGFR G A Citation Format: Maria del Mar Alvarez-Torres, Xi Fu, Romella K Sagatelian, Alejandro L Buendia, Raul Rabadan. Unveiling the functional impact of glioma non-coding germline variants using large language models [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4887.

Prioritizing cardiovascular disease-associated variants altering NKX2-5 and TBX5 binding through an integrative computational approach

Cardiovascular diseases (CVDs) are the leading cause of death worldwide and are heavily influenced by genetic factors. Genome-wide association studies (GWAS) have mapped > 90% of CVD-associated variants within the non-coding genome, which can alter the function of regulatory proteins, like transcription factors (TFs). However, due to the overwhelming number of GWAS single nucleotide polymorphisms (SNPs) (>500,000), prioritizing variants for in vitro analysis remains challenging. In this work, we implemented a computational approach that considers support vector machine (SVM)-based TF binding site classification and cardiac expression quantitative trait loci (eQTL) analysis to identify and prioritize potential CVD-causing SNPs. We identified 1,535 CVD-associated SNPs that occur within TF footprints and putative enhancers in the human heart and 14,218 variants in linkage disequilibrium (LD) with genotype-dependent gene expression in cardiac tissue. Using ChIP-seq data from two essential cardiac TFs (NKX2-5 and TBX5) in human induced pluripotent stem cell-derived cardiomyocytes (hIPSC-CM), we trained a large-scale gapped k-mer SVM (LS-GKM-SVM) model to identify CVD-associated SNPs that altered NKX2-5 and TBX5 binding sites. The model was tested by scoring human heart TF genomic footprints within putative enhancers and measuring in vitro binding through electrophoretic mobility shift assay (EMSA). Five variants predicted to alter NKX2-5 (rs59310144, rs6715570, and rs61872084) and TBX5 (rs7612445 and rs7790964) binding were prioritized for in vitro validation based on the magnitude of the predicted change in binding and because they are eQTLs in cardiac tissue. All five variants altered NKX2-5 and TBX5 DNA binding. In summary, we present a bioinformatic approach that considers tissue-specific eQTL analysis and SVM-based TF binding site classification to prioritize CVD-associated variants for in vitro experimental analysis.

Genetic evidence implicating natriuretic peptide receptor-3 in cardiovascular disease risk: a Mendelian randomization study

BackgroundC-type natriuretic peptide (CNP) is a known target for promoting growth and has been implicated as a therapeutic opportunity for the prevention and treatment of cardiovascular disease (CVD). This study aimed to explore the effect of CNP on CVD risk using the Mendelian randomization (MR) framework.MethodsInstrumental variables mimicking the effects of pharmacological intervention on CNP were identified as uncorrelated genetic variants located in the genes coding for its primary receptors, natriuretic peptide receptors-2 and 3 (NPR2 and NPR3), that associated with height. We performed MR and colocalization analyses to investigate the effects of NPR2 signalling and NPR3 function on CVD outcomes and risk factors. MR estimates were compared to those obtained when considering height variants from throughout the genome.ResultsGenetically-proxied reduced NPR3 function was associated with a lower risk of CVD, with odds ratio (OR) 0.74 per standard deviation (SD) higher NPR3-predicted height, and 95% confidence interval (95% CI) 0.64–0.86. This effect was greater in magnitude than observed when considering height variants from throughout the genome. For CVD subtypes, similar MR associations for NPR3-predicted height were observed when considering the outcomes of coronary artery disease (0.75, 95% CI 0.60–0.92), stroke (0.69, 95% CI 0.50–0.95) and heart failure (0.77, 95% CI 0.58–1.02). Consideration of CVD risk factors identified systolic blood pressure (SBP) as a potential mediator of the NPR3-related CVD risk lowering. For stroke, we found that the MR estimate for NPR3 was greater in magnitude than could be explained by a genetically predicted SBP effect alone. Colocalization results largely supported the MR findings, with no evidence of results being driven by effects due to variants in linkage disequilibrium. There was no MR evidence supporting effects of NPR2 on CVD risk, although this null finding could be attributable to fewer genetic variants being identified to instrument this target.ConclusionsThis genetic analysis supports the cardioprotective effects of pharmacologically inhibiting NPR3 receptor function, which is only partly mediated by an effect on blood pressure. There was unlikely sufficient statistical power to investigate the cardioprotective effects of NPR2 signalling.

Analysis of genetically determined gene expression suggests role of inflammatory processes in exfoliation syndrome

BackgroundExfoliation syndrome (XFS) is an age-related systemic disorder characterized by excessive production and progressive accumulation of abnormal extracellular material, with pathognomonic ocular manifestations. It is the most common cause of secondary glaucoma, resulting in widespread global blindness. The largest global meta-analysis of XFS in 123,457 multi-ethnic individuals from 24 countries identified seven loci with the strongest association signal in chr15q22–25 region near LOXL1. Expression analysis have so far correlated coding and a few non-coding variants in the region with LOXL1 expression levels, but functional effects of these variants is unclear. We hypothesize that analysis of the contribution of the genetically determined component of gene expression to XFS risk can provide a powerful method to elucidate potential roles of additional genes and clarify biology that underlie XFS.ResultsTranscriptomic Wide Association Studies (TWAS) using PrediXcan models trained in 48 GTEx tissues leveraging on results from the multi-ethnic and European ancestry GWAS were performed. To eliminate the possibility of false-positive results due to Linkage Disequilibrium (LD) contamination, we i) performed PrediXcan analysis in reduced models removing variants in LD with LOXL1 missense variants associated with XFS, and variants in LOXL1 models in both multiethnic and European ancestry individuals, ii) conducted conditional analysis of the significant signals in European ancestry individuals, and iii) filtered signals based on correlated gene expression, LD and shared eQTLs, iv) conducted expression validation analysis in human iris tissues. We observed twenty-eight genes in chr15q22–25 region that showed statistically significant associations, which were whittled down to ten genes after statistical validations. In experimental analysis, mRNA transcript levels for ARID3B, CD276, LOXL1, NEO1, SCAMP2, and UBL7 were significantly decreased in iris tissues from XFS patients compared to control samples. TWAS genes for XFS were significantly enriched for genes associated with inflammatory conditions. We also observed a higher incidence of XFS comorbidity with inflammatory and connective tissue diseases.ConclusionOur results implicate a role for connective tissues and inflammation pathways in the etiology of XFS. Targeting the inflammatory pathway may be a potential therapeutic option to reduce progression in XFS.

Metabolomic Investigation of Major Depressive Disorder Identifies a Potentially Causal Association With Polyunsaturated Fatty Acids

BackgroundMetabolic differences have been reported between individuals with and without major depressive disorder (MDD), but their consistency and causal relevance have been unclear. MethodsWe conducted a metabolome-wide association study of MDD with 249 metabolomic measures available in the UK Biobank (n = 29,757). We then applied two-sample bidirectional Mendelian randomization and colocalization analysis to identify potentially causal relationships between each metabolite and MDD. ResultsA total of 191 metabolites tested were significantly associated with MDD (false discovery rate–corrected p < .05), which decreased to 129 after adjustment for likely confounders. Lower abundance of omega-3 fatty acid measures and a higher omega-6 to omega-3 ratio showed potentially causal effects on liability to MDD. There was no evidence of a causal effect of MDD on metabolite levels. Furthermore, genetic signals associated with docosahexaenoic acid colocalized with loci associated with MDD within the fatty acid desaturase gene cluster. Post hoc Mendelian randomization of gene-transcript abundance within the fatty acid desaturase cluster demonstrated a potentially causal association with MDD. In contrast, colocalization analysis did not suggest a single causal variant for both transcript abundance and MDD liability, but rather the likely existence of two variants in linkage disequilibrium with one another. ConclusionsOur findings suggest that decreased docosahexaenoic acid and increased omega-6 to omega-3 fatty acids ratio may be causally related to MDD. These findings provide further support for the causal involvement of fatty acids in MDD.

Statistical learning for sparser fine-mapped polygenic models: The prediction of LDL-cholesterol.

Polygenic risk scores quantify the individual genetic predisposition regarding a particular trait. We propose and illustrate the application of existing statistical learning methods to derive sparser models for genome-wide data with a polygenic signal. Our approach is based on three consecutive steps. First, potentially informative loci are identified by a marginal screening approach. Then, fine-mapping is independently applied for blocks of variants in linkage disequilibrium, where informative variants are retrieved by using variable selection methods including boosting with probing and stochastic searches with the Adaptive Subspace method. Finally, joint prediction models with the selected variants are derived using statistical boosting. In contrast to alternative approaches relying on univariate summary statistics from genome-wide association studies, our three-step approach enables to select and fit multivariable regression models on large-scale genotype data. Based on UK Biobank data, we develop prediction models for LDL-cholesterol as a continuous trait. Additionally, we consider a recent scalable algorithm for the Lasso. Results show that statistical learning approaches based on fine-mapping of genetic signals result in a competitive prediction performance compared to classical polygenic risk approaches, while yielding sparser risk models.

Sparse block signal detection and identification for shared cross-trait association analysis

Genome-wide association studies (GWAS) have identified thousands of single nucleotide polymorphisms (SNPs) that are associated with complex traits. GWAS data allows us to investigate the shared genetic etiologies among different traits. However, linkage disequilibrium (LD) between the SNPs complicates the detection and identification of shared genetic effects. In this paper we model the LD by dividing the genome into LD blocks and linking the genetic variants within a block to a possible latent causal variant. An eigenvector-projected score statistic that leverages the set of variants in LD and a maxtype test statistic (Max-block) are proposed to detect the existence of cross-trait genetic association. The Max-block is easy to calculate and is shown to control the genome-wide error rate. After the detection a stepwise procedure is proposed to identify the significant blocks that explain the genetic sharing between two traits. Simulation experiments show that Max-block is more powerful than standard approaches in the sparse settings and is robust to different signal strengths or levels of sparsity. The method is applied to study shared cross-trait associations in 10 pediatric autoimmune diseases and identified several regions that explain the genetic sharing between juvenile idiopathic arthritis (JIA) and ulcerative colitis (UC) and between UC and Crohn’s disease (CD). In addition, our analysis also indicates the genetic sharing in the MHC region among systemic lupus (SLE), celiac disease (CEL) and common variable immunodeficiency (CVID). Results from real data and simulation studies show that Max-block provides an important alternative to commonly used genetic correlation estimation in understanding genetic correlation among complex diseases.

130 Does visual imagery vividness have a genetic basis? A genome-wide associa- tion study of 1019 individuals

The vividness of visual imagery varies widely between individuals and its extremes are associated with prosopagnosia and synaesthesia. A complete lack of visual imagery, recently termed aphantasia, is typically present from birth and is more common in those with affected family members. It is therefore postulated that there is a genetic component to the variation in visual imagery vividness. To try and identify genetic variants associated with visual imagery vividness we performed a genome-wide association study of 1019 individuals in the Exeter 10000 cohort. Participants who had previously been genotyped were then asked to complete the Vividness of Visual Imagery Questionnaire (VVIQ). We did not identify any variants that reached the genome-wide significance p&lt;5x10–8 level. However, 2 variants in linkage disequilibrium were below the p&lt;5x10–7 level and lie within an intron of the SYT1 gene which encodes a protein important in neuronal synaptic transmission. There were no significant genetic correlations with VVIQ and other publicly available traits. Further work should focus on using larger cohorts which will have greater power to identify variants which account for smaller amounts of VIVQ variance. This would also determine whether the SYT1 variant is truly significant.jacobday@doctors.org.uk

OA24 Predicting drug immunogenicity to tumour necrosis factor inhibitors in patients with rheumatoid arthritis

Abstract Background/Aims Rheumatoid arthritis (RA) is a chronic immune-mediated inflammatory disease that irreversibly damages synovial joints. Tumour necrosis factor inhibitors (TNFi) may be prescribed for patients with moderate to severe RA. However, approximately 40% of patients do not respond well to TNFi, with a lack of effectiveness underpinned by immunogenicity (development of anti-drug antibodies) in some patients. A recent well-powered study by Sazonovs et al. (2020) reported a strong genetic association between the HLA-DQA1 locus and immunogenicity to TNFi in a cohort of Crohn's disease patients treated with either infliximab or adalimumab. The current study aimed to investigate the association between HLA alleles, primarily HLA-DQA1, and immunogenicity to adalimumab in RA patients. Methods Anti-adalimumab antibody titres were measured using radioimmunoassay in serum samples from RA patients following six months of treatment. Genetic datasets were generated using the Illumina OmniExpress genotyping array and, following standard quality control filtering, variants were imputed using SNP2HLA statistical software. Following imputation, filtering was performed to exclude variants in linkage disequilibrium or with low minor allele frequency. Logistic regression models were developed to explore the association between HLA alleles and amino acids and immunogenicity (all models were adjusted for sex). Amino acid positions of interest were annotated using P2Rank; a ligand binding site prediction tool. Results A total of 171 HLA alleles and 339 amino acid positions were analysed in 226 RA patients. The strongest association observed between HLA alleles and immunogenicity was for HLA-DQB1*02 (coefficient = 0.87; 95% CI, 0.31-1.43; p value = 2.10E-3), followed by HLA-DQA1*05 (coefficient = 0.91; 95% CI, 0.29-1.53; p value = 3.95E-3). The strongest associated amino acid positions were amino acids 66 and 67 within the ligand binding domain of the HLA-DQB1 protein. Conclusion In keeping with previous findings in Crohn's disease, variants within HLA-DQ molecules confer the strongest evidence for immunogenicity risk in RA patients receiving adalimumab. RA patients carrying these variants could be redirected to receive etanercept, an effective alternative to monoclonal antibodies that is not immunogenic. Further well-powered studies in RA are now warranted to replicate these findings. Disclosure M. Alshehri: None. N. Nair: None. C. Yap: None. J.D. Isaacs: None. K. Hyrich: Honoraria; Abbvie. A.W. Morgan: None. A.G. Wilson: None. A. Barton: None. D. Plant: None.

Pklr Is a Genetic Modifier of Sickle Cell Disease

Background: Acute pain, the most prominent complication of sickle cell disease (SCD), results from vasoocclusion triggered by sickling of deoxygenated red blood cells (RBCs). A key factor influencing HbS oxygenation is the intracellular concentration of 2,3- diphosphoglycerate (2,3-DPG). 2,3-DPG, an intermediate substrate in the glycolytic pathway, decreases oxygen binding and stabilizes the deoxygenated hemoglobin. Pyruvate kinase (gene PKLR, protein PKR) is a rate-limiting enzyme in glycolysis; variants in PKLR may affect PKR activity, 2,3-DPG levels in RBCs, subsequent frequency of sickling and acute pain episodes (APE). There is thus a strong biological basis for exploring PKLR as a candidate gene affecting acute pain in SCD.Methods: The study population for genetic association consists of 2 cohorts: 1) 242 adults with HbSS from King's College Hospital (KCH), London, UK, with complete hospitalisation records over 10 years (2004-2013 inclusive) as the “discovery” cohort; 2) 977 children with HbSS or HbSb 0 thalassemia from the Silent Infarct Transfusion (SIT) trial, with a 3-year history of severe vasoocclusive pain based on hospitalization, as the “validation” cohort. Both studies were approved by the local Institutional Review Boards at KCH and Vanderbilt University Medical Center, respectively. An independent cohort comprises 52 adults with SCD enrolled under 3 protocols - NCT00011648, NCT00081523, and NCT03685721 - approved by the NHLBI Review Board (NIH), for evaluation of imbalance in allele expression.Genome scan for the KCH cohort was performed using llumina's Infinium “MEGA” chip (1.7m markers). The SIT DNA samples were genotyped using Illumina HumanHap650Y array 5 (661K markers) or Illumina Infinium HumanOmni1-Quad array (1.1m markers). The results were quality controlled followed by genotype imputation based on the 1000 Genomes Project phase 3 data. An annualised “hospitalisation rate” as a measure of pain incidence rate, was calculated by dividing the number of hospital admissions for severe acute pain by the number of years of observation for KCH and SIT cohorts (Fig A).We performed association analysis with common SNPs at PKLR locus using data from our genome-wide SNP set and a linear mixed modelling approach incorporating a genetic relatedness matrix to take account of relatedness, plus sex and age as fixed covariates.We corrected for multiple testing after quantifying the linkage disequilibrium (LD) within PKLR and used this to calculate appropriate significance levels. For the PKLR region and hospitalisation rate, the modified significance level was p<0.001268 for the discovery (KCH) cohort. For the allele expression assays, a synonymous variant, rs1052176 (R596R), in exon 11 of PKLR acted as a marker of relative expression levels of the 2 alleles of the gene. Allele specific expression was carried using the Bio-Rad digital droplet PCR system.Results: 7 of 47 variants evaluated in PKLR were associated with hospitalization rate (LnLnHospRate) in the discovery cohort: intron 4 - rs071053, and intron 2 - rs8177970, rs116244351, rs114455416, rs12741350, rs3020781, and rs8177964). All 7 were validated in Fisher's meta-analyses of the KCH and the SIT cohorts using p<0.0071 as threshold to correct for multiple testing (Fig B). We examined the pairwise LD between PKLR variants, and found all the intron 2 variants in tight LD, while R596R belongs to another LD block (Fig C). 52 SCD individuals had the R596R variant, of which 29 were heterozygous and 23 homozygous for the intron 2 haplotype associated with APE in SCD. We performed a Wilcoxon rank sum test and compared the variation in PKLR expression between the 2 alleles in subjects homozygous and heterozygous for the wildtype intron 2 haplotype, using genomic DNA as internal control for each subject. The results reveal a significant deviation from the expected expression ratio in those heterozygous for the intron 2 haplotype (mean 0.2073, +/- SD 0.0135) when compared with to those without the variant (mean 0.1239, +/- SD 0.0682), p=0.0297 (Fig D).Conclusion: Intronic variants of PKLR are associated with hospitalization rate for acute pain episodes in adults and children with SCD. We show that the intronic variants are likely to influence acute pain by affecting expression of the PKLR gene using allele-specific expression analyses, although the causal variant is unclear. These results support PKLR as a genetic modifier of SCD. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.

Chronic Pancreatitis: The True Pathogenic Culprit within the SPINK1 N34S-Containing Haplotype Is No Longer at Large.

A diverse range of loss-of-function variants in the SPINK1 gene (encoding pancreatic secretory trypsin inhibitor) has been identified in patients with chronic pancreatitis (CP). The haplotype harboring the SPINK1 c.101A>G (p.Asn34Ser or N34S) variant (rs17107315:T>C) is one of the most important heritable risk factors for CP as a consequence of its relatively high prevalence worldwide (population allele frequency ≈ 1%) and its considerable effect size (odds ratio ≈ 11). The causal variant responsible for this haplotype has been intensively investigated over the past two decades. The different hypotheses tested addressed whether the N34S missense variant has a direct impact on enzyme structure and function, whether c.101A>G could affect pre-mRNA splicing or mRNA stability, and whether another variant in linkage disequilibrium with c.101A>G might be responsible for the observed association with CP. Having reviewed the currently available genetic and experimental data, we conclude that c.-4141G>T (rs142703147:C>A), which disrupts a PTF1L-binding site within an evolutionarily conserved HNF1A-PTF1L cis-regulatory module located ∼4 kb upstream of the SPINK1 promoter, can be designated as the causal variant beyond reasonable doubt. This case illustrates the difficulties inherent in determining the identity of the causal variant underlying an initially identified disease association.

GWAS-Top Polymorphisms Associated With Late-Onset Alzheimer Disease in Brazil: Pointing Out Possible New Culprits Among Non-Coding RNAs.

Several genome-wide association studies (GWAS) have been carried out with late-onset Alzheimer’s disease (LOAD), mainly in European and Asian populations. Different polymorphisms were associated, but several of them without a functional explanation. GWAS are fundamental for identifying loci associated with diseases, although they often do not point to causal polymorphisms. In this sense, functional investigations are a fundamental tool for discovering causality, although the failure of this validation does not necessarily indicate a non-causality. Furthermore, the allele frequency of associated genetic variants may vary widely between populations, requiring replication of these associations in other ethnicities. In this sense, our study sought to replicate in 150 AD patients and 114 elderly controls from the South Brazilian population 18 single-nucleotide polymorphisms (SNPs) associated with AD in European GWAS, with further functional investigation using bioinformatic tools for the associated SNPs. Of the 18 SNPs investigated, only four were associated in our population: rs769449 (APOE), rs10838725 (CELF1), rs6733839, and rs744373 (BIN1–CYP27C1). We identified 54 variants in linkage disequilibrium (LD) with the associated SNPs, most of which act as expression or splicing quantitative trait loci (eQTLs/sQTLs) in genes previously associated with AD or with a possible functional role in the disease, such as CELF1, MADD, MYBPC3, NR1H3, NUP160, SPI1, and TOMM40. Interestingly, eight of these variants are located within long non-coding RNA (lncRNA) genes that have not been previously investigated regarding AD. Some of these polymorphisms can result in changes in these lncRNAs’ secondary structures, leading to either loss or gain of microRNA (miRNA)-binding sites, deregulating downstream pathways. Our pioneering work not only replicated LOAD association with polymorphisms not yet associated in the Brazilian population but also identified six possible lncRNAs that may interfere in LOAD development. The results lead us to emphasize the importance of functional exploration of associations found in large-scale association studies in different populations to base personalized and inclusive medicine in the future.

Estimating colocalization probability from limited summary statistics

BackgroundColocalization is a statistical method used in genetics to determine whether the same variant is causal for multiple phenotypes, for example, complex traits and gene expression. It provides stronger mechanistic evidence than shared significance, which can be produced through separate causal variants in linkage disequilibrium. Current colocalization methods require full summary statistics for both traits, limiting their use with the majority of reported GWAS associations (e.g. GWAS Catalog). We propose a new approximation to the popular coloc method that can be applied when limited summary statistics are available.Our method (POint EstiMation of Colocalization, POEMColoc) imputes missing summary statistics for one or both traits using LD structure in a reference panel, and performs colocalization using the imputed summary statistics.ResultsWe evaluate the performance of POEMColoc using real (UK Biobank phenotypes and GTEx eQTL) and simulated datasets. We show good correlation between posterior probabilities of colocalization computed from imputed and observed datasets and similar accuracy in simulation. We evaluate scenarios that might reduce performance and show that multiple independent causal variants in a region and imputation from a limited subset of typed variants have a larger effect while mismatched ancestry in the reference panel has a modest effect. Further, we find that POEMColoc is a better approximation of coloc when the imputed association statistics are from a well powered study (e.g., relatively larger sample size or effect size). Applying POEMColoc to estimate colocalization of GWAS Catalog entries and GTEx eQTL, we find evidence for colocalization of 150,000 trait-gene-tissue triplets.ConclusionsWe find that colocalization analysis performed with full summary statistics can be closely approximated when only the summary statistics of the top SNP are available for one or both traits. When applied to the full GWAS Catalog and GTEx eQTL, we find that colocalized trait-gene pairs are enriched in tissues relevant to disease etiology and for matches to approved drug mechanisms. POEMColoc R package is available at https://github.com/AbbVie-ComputationalGenomics/POEMColoc.

Variant Intronic Enhancer Controls SCN10A-short Expression and Heart Conduction.

Genetic variants in SCN10A, encoding the neuronal voltage-gated sodium channel NaV1.8, are strongly associated with atrial fibrillation, Brugada syndrome, cardiac conduction velocities, and heart rate. The cardiac function of SCN10A has not been resolved, however, and diverging mechanisms have been proposed. Here, we investigated the cardiac expression of SCN10A and the function of a variant-sensitive intronic enhancer previously linked to the regulation of SCN5A, encoding the major essential cardiac sodium channel NaV1.5. The expression of SCN10A was investigated in mouse and human hearts. With the use of CRISPR/Cas9 genome editing, the mouse intronic enhancer was disrupted, and mutant mice were characterized by transcriptomic and electrophysiological analyses. The association of genetic variants at SCN5A-SCN10A enhancer regions and gene expression were evaluated by genome-wide association studies single-nucleotide polymorphism mapping and expression quantitative trait loci analysis. We found that cardiomyocytes of the atria, sinoatrial node, and ventricular conduction system express a short transcript comprising the last 7 exons of the gene (Scn10a-short). Transcription occurs from an intronic enhancer-promoter complex, whereas full-length Scn10a transcript was undetectable in the human and mouse heart. Expression quantitative trait loci analysis revealed that the genetic variants in linkage disequilibrium with genetic variant rs6801957 in the intronic enhancer associate with SCN10A transcript levels in the heart. Genetic modification of the enhancer in the mouse genome led to reduced cardiac Scn10a-short expression in atria and ventricles, reduced cardiac sodium current in atrial cardiomyocytes, atrial conduction slowing and arrhythmia, whereas the expression of Scn5a, the presumed enhancer target gene, remained unaffected. In patch-clamp transfection experiments, expression of Scn10a-short-encoded NaV1.8-short increased NaV1.5-mediated sodium current. We propose that noncoding genetic variation modulates transcriptional regulation of Scn10a-short in cardiomyocytes that impacts NaV1.5-mediated sodium current and heart rhythm. Genetic variants in and around SCN10A modulate enhancer function and expression of a cardiac-specific SCN10A-short transcript. We propose that noncoding genetic variation modulates transcriptional regulation of a functional C-terminal portion of NaV1.8 in cardiomyocytes that impacts on NaV1.5 function, cardiac conduction velocities, and arrhythmia susceptibility.

Germline variants at SOHLH2 influence multiple myeloma risk

Multiple myeloma (MM) is caused by the uncontrolled, clonal expansion of plasma cells. While there is epidemiological evidence for inherited susceptibility, the molecular basis remains incompletely understood. We report a genome-wide association study totalling 5,320 cases and 422,289 controls from four Nordic populations, and find a novel MM risk variant at SOHLH2 at 13q13.3 (risk allele frequency = 3.5%; odds ratio = 1.38; P = 2.2 × 10−14). This gene encodes a transcription factor involved in gametogenesis that is normally only weakly expressed in plasma cells. The association is represented by 14 variants in linkage disequilibrium. Among these, rs75712673 maps to a genomic region with open chromatin in plasma cells, and upregulates SOHLH2 in this cell type. Moreover, rs75712673 influences transcriptional activity in luciferase assays, and shows a chromatin looping interaction with the SOHLH2 promoter. Our work provides novel insight into MM susceptibility.

Haplotype structure defines effects of common DPYD variants c.85T > C (rs1801265) and c.496A > G (rs2297595) on dihydropyrimidine dehydrogenase activity: Implication for 5-fluorouracil toxicity.

AimsThe aim of this study was to identify risk variants and haplotypes that impair dihydropyrimidine dehydrogenase (DPD) activity and are, therefore, candidate risk variants for severe toxicity to 5‐fluorouracil (5‐FU) chemotherapy.MethodsPlasma dihydrouracil/uracil (UH2/U) ratios were measured as a population marker for DPD activity in a total of 1382 subjects from 4 independent studies. Genotype and haplotype correlations with UH2/U ratios were assessed.ResultsSignificantly lower UH2/U ratios (panova < 2 × 10−16) were observed in carriers of the 4 well‐studied 5‐FU toxicity risk variants with mean differences (MD) of −43.7% for DPYD c.1905 + 1G > A (rs3918290), −46.0% for DPYD c.1679T > G (rs55886062), −37.1%, for DPYD c.2846A > T (rs67376798), and −13.2% for DPYD c.1129‐5923C > G (rs75017182). An additional variant, DPYD c.496A > G (rs2297595), was also associated with lower UH2/U ratios (P < .0001, MD: −12.6%). A haplotype analysis was performed for variants in linkage disequilibrium with c.496A > G, which consisted of the common variant c.85T > C (rs1801265) and the risk variant c.1129‐5923C > G. Both haplotypes carrying c.496A > G were associated with decreased UH2/U ratios (H3, P = .003, MD: −9.6%; H5, P = .002, MD: −16.9%). A haplotype carrying only the variant c.85T > C (H2) was associated with elevated ratios (P = .004, MD: +8.6%).ConclusionsBased on our data, DPYD‐c.496A > G is a strong candidate risk allele for 5‐FU toxicity. Our data suggest that DPYD‐c.85T > C might be protective; however, the deleterious impacts of the linked alleles c.496A > G and c.1129‐5923C > G likely limit this effect in patients. The possible protective effect of c.85T > C and linkage disequilibrium with c.496A > G and c.1129‐5923C > G may have hampered prior association studies and should be considered in future clinical studies.

A linkage and exome study of multiplex families with bipolar disorder implicates rare coding variants of ANK3 and additional rare alleles at 10q11-q21.

BackgroundBipolar disorder is a highly heritable psychiatric condition for which specific genetic factors remain largely unknown. In the present study, we used combined whole-exome sequencing and linkage analysis to identify risk loci and dissect the contribution of common and rare variants in families with a high density of illness.MethodsOverall, 117 participants from 15 Australian extended families with bipolar disorder (72 with affective disorder, including 50 with bipolar disorder type I or II, 13 with schizoaffective disorder–manic type and 9 with recurrent unipolar disorder) underwent whole-exome sequencing. We performed genome-wide linkage analysis using MERLIN and conditional linkage analysis using LAMP. We assessed the contribution of potentially functional rare variants using a gene-based segregation test.ResultsWe identified a significant linkage peak on chromosome 10q11-q21 (maximal single nucleotide polymorphism = rs10761725; exponential logarithm of the odds [LODexp] = 3.03; empirical p = 0.046). The linkage interval spanned 36 protein-coding genes, including a gene associated with bipolar disorder, ankyrin 3 (ANK3). Conditional linkage analysis showed that common ANK3 risk variants previously identified in genome-wide association studies — or variants in linkage disequilibrium with those variants — did not explain the linkage signal (rs10994397 LOD = 0.63; rs9804190 LOD = 0.04). A family-based segregation test with 34 rare variants from 14 genes under the linkage interval suggested rare variant contributions of 3 brain-expressed genes: NRBF2 (p = 0.005), PCDH15 (p = 0.002) and ANK3 (p = 0.014).LimitationsWe did not examine non-coding variants, but they may explain the remaining linkage signal.ConclusionCombining family-based linkage analysis with next-generation sequencing data is effective for identifying putative disease genes and specific risk variants in complex disorders. We identified rare missense variants in ANK3, PCDH15 and NRBF2 that could confer disease risk, providing valuable targets for functional characterization.

Genome‐wide profiling of the noncoding regulatory mechanisms in Alzheimer’s disease

AbstractBackgroundGenome‐wide association studies (GWAS) have identified genetic variants associated with Alzheimer’s Disease (AD), but the majority are noncoding and may affect transcriptional regulatory elements such as enhancers. While prior post‐GWAS analyses focused on analyzing only top GWAS signals, this study focuses on a systematic genome‐wide analysis of all significant GWAS signals aimed at identifying a more complete set of causal variants given a complex linkage disequilibrium (LD) genome structures and the regulatory mechanisms, and tissues/cell types.MethodTo perform genome‐wide analysis of GWAS summary statistics (NIAGADS NG00036) from the IGAP study [Lambert et al., 2013], we adopted INFERNO (http://inferno.lisanwanglab.org), a recent method for post‐GWAS analysis of non‐coding variants. INFERNO integrates GWAS data with hundreds of publicly available functional genomics datasets (e.g., ENCODE, FANTOM5, GTEx, HOMER) across 485 tissues and cell types to facilitate identification of the regulatory mechanisms and tissue/cell types affected by noncoding genetic variants.ResultsGenome‐wide INFERNO analysis identified 168 distinct loci by pruning all genome‐wide significant (p‐value &lt; 5e‐8) GWAS variants using LD structure, a significant increase compared to original top‐SNP analysis (19 loci). Across these 168 distinct loci, INFERNO identified 1,358 candidate potentially causal variants in LD. Around 6% of these variants were localized in coding exons, 67% overlapped enhancers in at least one tissue and 57% overlapped transcription factor binding sites. Co‐localization analysis of GWAS and eQTL data (GTEx_v6p) identified tissue‐specific effects of 53 variants on 43 genes across 16 INFERNO tissue categories, with 12 variants affecting expression of 16 genes in the blood (RP11‐1036E20.9, ZNF223, ZNF285B, RP11‐286H15.1, FAM115D, EPHA1‐AS1, TAS2R60, TAS2R62P, PTK2B) and brain (CR1, ABCA7, POLR2E, EML2, PPM1N, RP11‐385F7.1, FAM115C). These INFERNO‐identified genes include both novel and genes previously implicated in AD. E.g. EPHA1‐AS1 activity was confirmed by luciferase reporter assay [Amlie‐Wolf et al., 2019]. Additionally, our analysis found that PTK2B gene, an early marker of Tau pathology [Dourlen et al., 2016], is regulated by rs28834970 that also disrupts binding motif of HLF(bZIP) in blood.ConclusionGenome‐wide INFERNO analysis confirmed experimentally validated and discovered additional plausible biological mechanisms related to AD. These can help to inform the development of better AD therapeutic targets.

Qtlizer: comprehensive QTL annotation of GWAS results

Exploration of genetic variant-to-gene relationships by quantitative trait loci such as expression QTLs is a frequently used tool in genome-wide association studies. However, the wide range of public QTL databases and the lack of batch annotation features complicate a comprehensive annotation of GWAS results. In this work, we introduce the tool “Qtlizer” for annotating lists of variants in human with associated changes in gene expression and protein abundance using an integrated database of published QTLs. Features include incorporation of variants in linkage disequilibrium and reverse search by gene names. Analyzing the database for base pair distances between best significant eQTLs and their affected genes suggests that the commonly used cis-distance limit of 1,000,000 base pairs might be too restrictive, implicating a substantial amount of wrongly and yet undetected eQTLs. We also ranked genes with respect to the maximum number of tissue-specific eQTL studies in which a most significant eQTL signal was consistent. For the top 100 genes we observed the strongest enrichment with housekeeping genes (P = 2 × 10–6) and with the 10% highest expressed genes (P = 0.005) after grouping eQTLs by r2 > 0.95, underlining the relevance of LD information in eQTL analyses. Qtlizer can be accessed via https://genehopper.de/qtlizer or by using the respective Bioconductor R-package (https://doi.org/10.18129/B9.bioc.Qtlizer).