Genome-wide Association Research Articles

The effect of depression status on osteoarthritis: A powerful two-step Mendelian randomization study.

There is no conclusive epidemiological evidence regarding the relationship between OA, depression, and whole-body fat mass. In this study, we conducted a two-step Mendelian randomization analysis to determine the causal relationships between them. The published summary-level data are from genome-wide association studies (GWAS). Our study included 357,957 samples and 10,828,862 SNPs. Finally, the outcome GWAS data for OA came from a GWAS on the genetic architecture of OA using UK Biobank data. This study included 50,508 samples and 15,845,511 SNPs. We used five different modes of analysis, including inverse variance weighted meta-analysis (IVW), MR-Egger regression, weighted median, simple mode, and weighted mode, to explore causal relationships. We found a positive correlation between depression and body fat mass, with depression leading to body fat mass an increase in (IVW result: p=3.39E-07, OR (95% CI) =2.16 (1.61, 2.90)). We also found a positive correlation between body fat mass and OA, with body fat mass increasing the risk of OA (IVW result: p=1.65E-33, OR (95% CI)=1.98 (1.77, 2.21). Body fat mass played an important role as a mediator in the causal relationship between depression and OA, with approximately 14% of the risk of OA caused by depression being mediated by body fat mass. Our study offers reliable evidence that depression has a detrimental impact on the risk of OA. Future research can support these associations from improving depressed effect, including social, biological, and behavioral factors, to reduce the risk of chronic diseases such as osteoarthritis. And we identified high-risk variation of alleles which associated with OA and depression can be used to predict disease and provide a basis for clinical intervention and treatment of OA.

Artificial intelligence-enabled electrocardiogram for mortality and cardiovascular risk estimation: a model development and validation study

Artificial intelligence (AI)-enabled electrocardiography (ECG) can be used to predict risk of future disease and mortality but has not yet been adopted into clinical practice. Existing model predictions do not have actionability at an individual patient level, explainability, or biological plausibi. We sought to address these limitations of previous AI-ECG approaches by developing the AI-ECG risk estimator (AIRE) platform. The AIRE platform was developed in a secondary care dataset (Beth Israel Deaconess Medical Center [BIDMC]) of 1 163 401 ECGs from 189 539 patients with deep learning and a discrete-time survival model to create a patient-specific survival curve with a single ECG. Therefore, AIRE predicts not only risk of mortality, but also time-to-mortality. AIRE was validated in five diverse, transnational cohorts from the USA, Brazil, and the UK (UK Biobank [UKB]), including volunteers, primary care patients, and secondary care patients. AIRE accurately predicts risk of all-cause mortality (BIDMC C-index 0·775, 95% CI 0·773-0·776; C-indices on external validation datasets 0·638-0·773), future ventricular arrhythmia (BIDMC C-index 0·760, 95% CI 0·756-0·763; UKB C-index 0·719, 95% CI 0·635-0·803), future atherosclerotic cardiovascular disease (0·696, 0·694-0·698; 0·643, 0·624-0·662), and future heart failure (0·787, 0·785-0·789; 0·768, 0·733-0·802). Through phenome-wide and genome-wide association studies, we identified candidate biological pathways for the prediction of increased risk, including changes in cardiac structure and function, and genes associated with cardiac structure, biological ageing, and metabolic syndrome. AIRE is an actionable, explainable, and biologically plausible AI-ECG risk estimation platform that has the potential for use worldwide across a wide range of clinical contexts for short-term and long-term risk estimation. British Heart Foundation, National Institute for Health and Care Research, and Medical Research Council.

Investigating the causal association between heme oxygenase-1 and asthma: A bidirectional two-sample Mendelian randomization analysis in a European population

BackgroundThe association between heme oxygenase-1 (HO-1) and asthma has been a subject of debate in both observational and experimental studies. We aimed to evaluate the potential causal relationship between HO-1 and asthma. Materials and methodsA bidirectional two-sample Mendelian randomization (TSMR) study was conducted to examine the causal relationship between HO-1 and asthma. In the forward Mendelian randomization (MR) analyses, HO-1 was considered as the exposure, while asthma as the outcome. Conversely, in the reverse MR analyses, asthma was regarded as the exposure, and HO-1 as the outcome. Data for HO-1 and asthma were obtained from publicly accessible genome-wide association studies (GWAS). These causal relationships were identified through 5 MR methods, namely MR-Egger, weighted median, inverse-variance weighted (IVW), simple mode, and weighted mode. Additionally, sensitivity tests were conducted to assess the robustness of MR study. Finally, additional asthma datasets and childhood asthma were selected to validate the findings. ResultsIn the forward MR analyses, according to the IVW method, genetically predicted HO-1 displays a negative correlation with the risk of asthma (OR 0.947, 95% CI 0.905–0.990). It was not found any SNP overly sensitive or disproportionately responsible for the outcome. No evidence of heterogeneity and pleiotropy between SNPs was observed. Genetically predicted asthma was not associated with HO-1 in reverse MR analyses using the IVW method. The same results were validated in additional asthma datasets and in childhood asthma. ConclusionThe results of MR analysis revealed heme oxygenase-1 as a protective factor for asthma.

The Causal Role of Thyroid Hormones in Bipolar Disorders: A Two-Sample Mendelian Randomization Study.

Bipolar disorder is a complex psychiatric condition with distinctions between clinical subtypes including Type 1 and 2 disorders. Several studies have proposed that thyroid hormones may be involved in the aetiology of bipolar disorders. This study employed a two-sample Mendelian randomization (MR) approach to investigate the causal relationships between six thyroid hormone metrics (TSH, FT4, FT3, TT3, FT3/FT4 and TT3/FT4) and bipolar disorder and Type 1 and 2 disorders, separately. Genome-wide association (GWAS) data from the ThyroidOmics Consortium (up to 271,040 individuals of European ancestry) were used for thyroid function metrics. Bipolar disorder GWAS data included 41,917 cases and 371,549 controls (25,060 Type 1 and 6,781 Type 2 cases). We applied inverse variance weighted (IVW) methods for primary MR analysis, with MR Egger, weighted median and weighted mode for sensitivity. Additional tests assessed horizontal pleiotropy and heterogeneity. Higher FT4 levels showed a protective causal effect against bipolar disorder (OR: 0.92, 95% CI: 0.86-0.97, p = 4.58 × 10-3) and a suggestive effect on Type 1 disorders (OR: 0.92, 95% CI: 0.86-0.99, p = 3.21 × 10-2). Elevated FT3 (OR: 1.18, 95% CI: 1.03-1.35, p = 1.55 × 10-2) and FT3/FT4 ratio (OR: 1.97, 95% CI: 1.02-3.82, p = 4.46 × 10-2) had suggestive harmful effects on Type 1 disorders. Sensitivity analyses showed consistent effects, with no significant horizontal pleiotropy or heterogeneity. These findings highlight the protective role of FT4 and the potentially harmful effect of elevated FT3 in Type 1 bipolar disorder, highlighting the need for further research on thyroid hormone levels as a potential treatment strategy for Type 1 bipolar disorder.

Genetically Predicted Sleep Traits and Sensorineural Hearing Loss: A Mendelian Randomization Study.

Observational studies suggest a potential association between sleep characteristics, sensorineural hearing loss (SNHL), and sudden SNHL (SSNHL), but causal evidence is scarce. We sought to clarify this issue using two-sample Mendelian randomization analysis. The inverse-variance weighted (IVW) method was performed as primary analysis to assess bidirectional causal associations between sleep traits (chronotype, sleep duration, insomnia, daytime sleepiness, and snoring) and SNHL/SSNHL using publicly available Genome-Wide Association Studies summary data from two large consortia (UK Biobank and FinnGen). Sensitivity analyses, including Mendelian randomization (MR)-Egger, Mendelian randomization pleiotropy residual sum and outlier, weight median, Cochran's Q test, leave-one-out analysis, and potential pleiotropy analysis, were conducted to ensure robustness. IVW analysis found suggestive associations of morning chronotype (odds ratio [OR] = 1.08, 95% confidence interval [CI] = 1.01-1.16, p = 0.031) and daytime sleepiness (OR = 1.88, 95% CI = 1.24-2.87, p = 0.003) with SNHL onset. Additionally, morning chronotype was nominally associated with SSNHL onset using IVW method (OR = 1.37, 95% CI = 1.10-1.71, p = 0.006). However, there was no evidence for the causal effect of SNHL and SSNHL on different sleep traits (all p > 0.05). Sensitivity analysis showed that the results were stable. Within the MR limitations, morning chronotype and daytime sleepiness were underlying causal contributors to the burden of SNHL, indicating that optimal sleep might facilitate the prevention and development of SNHL. 3 Laryngoscope, 134:4723-4729, 2024.

Discover QTLs for the level of blood components in Shaoxing duck using GWAS and haplotype sharing analysis

Blood composition is indicative of health-related traits such as immunity and metabolism. The use of molecular genetics to investigate alterations in these attributes in laying ducks is a novel approach. Our objective was to employ genome − wide association studies (GWAS) and haplotype − sharing analysis to identify genomic regions and potential genes associated with 11 blood components in Shaoxing ducks. Our findings revealed 35 SNPs and 1 SNP associated with low-density lipoprotein cholesterol (LDL) and globulin (GLB), respectively. We identified 36 putative candidate genes for the LDL trait in close proximity to major QTLs and key loci. Based on their biochemical and physiological properties, TRA2A, NPY, ARHGEF26, DHX36, and AADAC are the strongest putative candidate genes. Through linkage disequilibrium analysis and haplotype sharing analysis, we identified three haplotypes and one haplotype, respectively, that were significantly linked with LDL and GLB. These haplotypes could be selected as potential candidate haplotypes for molecular breeding of Shaoxing ducks. Additionally, we utilized a bootstrap test to verify the reliability of GWAS with small experimental samples. The test can be accessed at https://github.com/xuwenwu24/Bootstrap-test.

Longitudinal Association of Changes in Metabolic Syndrome with Cognitive Function: 12-Year Follow-up of the Guangzhou Biobank Cohort Study.

The association of changes in metabolic syndrome (MetS) with cognitive function remains unclear. We explored this association using prospective and Mendelian randomization (MR) studies. MetS components including high-density lipoprotein cholesterol (HDL-C), systolic blood pressure (SBP), waist circumference (WC), fasting plasma glucose (FPG), and triglycerides were measured at baseline and two follow-ups, constructing a MetS index. Immediate, delayed memory recall, and cognitive function along with its dimensions were assessed by immediate 10- word recall test (IWRT) and delayed 10-word recall test (DWRT), and mini-mental state examination (MMSE), respectively, at baseline and follow-ups. Linear mixed-effect model was used. Additionally, the genome-wide association study (GWAS) of MetS was conducted and one-sample MR was performed to assess the causality between MetS and cognitive function. Elevated MetS index was associated with decreasing annual change rates (decrease) in DWRT and MMSE scores, and with decreases in attention, calculation and recall dimensions. HDL-C was positively associated with an increase in DWRT scores, while SBP and FPG were negatively associated. HDL-C showed a positive association, whereas WC was negatively associated with increases in MMSE scores, including attention, calculation and recall dimensions. Interaction analysis indicated that the association of MetS index on cognitive decline was predominantly observed in low family income group. The GWAS of MetS identified some genetic variants. MR results showed a non-significant causality between MetS and decrease in DWRT, IWRT, nor MMSE scores. Our study indicated a significant association of MetS and its components with declines in memory and cognitive function, especially in delayed memory recall.

Ancestry and genome-wide association study of domestic pigs that survive African swine fever in Uganda

African swine fever (ASF) is endemic to Uganda and causes annual outbreaks. Some pigs survive these outbreaks and remain asymptomatic but are African swine fever virus (ASFV) positive. The potential heritability and genetic disparities in disease susceptibility among Ugandan pigs are not fully understood. In a 12-year study, whole blood and tissue samples were collected from 212 pigs across 19 districts in Uganda. Polymerase chain reaction (PCR) assays were used to determine ASFV infection status and genotyping was completed using a commercial porcine array. The point prevalence of ASF was calculated for each district, and breed composition origins were quantified for the sampled pigs by implementing established ancestry analyses. Genome-wide associated studies (GWAS) were conducted using all available domestic swine samples (full study population; n = 206) as well as a reduced dataset (farm-level study population; n = 129). This study revealed a greater number of ASFV-positive pigs in border districts than in non-border districts, a high level of admixture among domestic pigs sampled from Ugandan smallholder farms, and 48 loci that were associated with ASFV infection status. The discovery of 48 significant SNPs and 28 putative candidate genes may imply the possibility of heritability for resistance to ASFV. However, additional investigations in ASFV-endemic regions are required to fully elucidate the heritability of ASFV susceptibility among surviving pigs in Uganda.

Genomic-Wide Association Studies For Resistance to Anthracnose in a Pseudo-F2 Mango Population

Goal: This study aimed to apply genome-wide association studies (GWAS) to identify single nucleotide polymorphisms (SNPs) associated with resistance to mango anthracnose. Theoretical Framework: The statistical methods of GWAS aim to identify associations among genotypes with phenotypes for marker-assisted selection (MAS) or genome dissection of important traits. Method: Ninety-four plants of a ‘Haden’ × ‘Tommy Atkins’ pseudo-F2 population were genotyped with 705 SNPs and phenotyped for resistance to Colletotrichum siamense, anthracnose causal agent, by spraying young leaves, three times at different periods, with 103 conidia/mL. Qualitative GWAS, with Plink Fisher’s exact test and Cochran-Armitage trend genotypic test, and quantitative GWAS, with Tassel general (GLM) and mixed linear models (MLM) were applied. F2 plants with symptoms mean absence >75% were considered moderately resistant to the fungus. Results and discussion: Qualitative and quantitative GWAS methods combined enable the identification of two consensus SNP loci controlling 24.76% of total variance to mango anthracnose resistance. Other four SNPs, identified, exclusively in quantitative or qualitative analysis, explained 28.55% of total variance to mango anthracnose. These SNPs loci are promising candidate for further PCR primers design in order to apply MAS in mango. Research implications: This study facilitates the identification of consensus chromosomal regions associated with mango anthracnose resistance. The application of MAS in mango breeding will accelerate the development of new cultivars. Originality/Value: To our knowledge, this a pioneering GWAS study applied to anthracnose, which is the most significant and widespread fungal disease affecting mango fruit globally.

Genome-wide association study of CMR image-based aortic stiffness in 45,789 individuals identifies loci linked with cardiovascular diseases

Abstract Background The use of artificial intelligence on large-scale cardiac magnetic resonance (CMR) imaging data has enabled detailed characterisation of cardiac shape and function for use in genome-wide association studies (GWAS), leading to improved understanding of genetic aetiology of cardiovascular diseases (CVDs). Aortic stiffness is associated with increased risk of cardiovascular events in observational studies. Here, we present a novel CMR-based pressure-independent measure of aortic stiffness, identify genetic associations and explore links with CVDs. Purpose Identify genetic contributions to pressure-independent aortic stiffness and links to CVDs using large-scale GWAS analysis on the UK Biobank cohort. Methods A pre-trained neural network was used to segment CMR images of participants in the UK Biobank to quantify ascending aorta diameter, which together with contemporaneous systolic and diastolic blood pressure measurements (SBP, DBP, respectively) were used to calculate a pressure-independent measure of aortic stiffness (β0). The measurement assumes an exponential relationship between pressure and aortic diameter. We then performed a GWAS on β0 in 45,789 participants of European ancestry without CVDs. A multi-trait-based conditional & joint analysis (mtCOJO) was performed to estimate genetic effects independent of SBP, DBP and aortic area. We mapped the independent lead variants associated with aortic stiffness at P < 5x10-8 to nearest gene, and characterised shared genetic association across CVDs using Open Targets Platform database. Results The GWAS identified 17 independent lead variants, including 6 that remained significant in the conditional analysis. Four of these lead variants, not previously reported in GWAS of other aortic traits, were located near CTD-2203A3.1, MCF2L, CFDP1 and CDH13 genes (red annotations in the Figure). MCF2L, CFDP1 and CDH13 have been associated with coronary artery disease, and CFDP1 with myocardial infarction and coronary atherosclerosis in previous studies. Conclusion A GWAS of CMR-derived aortic stiffness identified 4 loci distinct from other aortic traits that also associate with other CVDs, suggesting genetic links between aortic stiffness and CVDs. Further analyses are required to explore biological mechanisms underlying these genetic associations.Aortic Stiffness Manhattan Plot

Does PhenoAge acceleration impact adverse outcomes following major surgery (rehospitalization or death)? A genome-wide association study and mendelian randomisation

Abstract Background Biological aging reflects the state of an individual's cellular and molecular function, and is distinct from chronological age; which reflects the passage of time[1]. Validated measures of biological aging, such as Phenotypic aging (PhenoAge) estimated from nine routine clinical biomarkers, have been associated with increased risk of age-related diseases and mortality in cohort studies[2]. However, the potential causal relationship between them remains unclear. In this study, we aim to identify genetic variants associated with PhenoAge Acceleration (PhenoAgeAccl) and adverse outcome after surgery (rehospitalisation or death within 365-days] and assess the causal relationship between these two phenotypes using Mendelian Randomization framework. Objectives -Perform a genome-wide association study(GWAS) on adverse outcomes following major surgery. -Assess the potential causal relationship between PhenoAgeAccl and adverse outcomes following major surgery using Mendelian randomization analysis. Methods We conducted a GWAS via plink using participants from the UKBiobank (Application 77596) to generate summary statistics for 6509 participants who had undergone major surgery within 365 days of their baseline assessment date(3). For quality control, SNPs were excluded if they didn't meet the following criteria: (1) imputation information score <0.3 (2) minor allele frequency <1%, (3) Hardy–Weinberg equilibrium test p-value significant at the Bonferroni-corrected level. Summary statistics for PhenoAgeAccl was obtained from a previous GWAS of European descents[4]. Gene enrichment analysis was performed using Multi-marker Analysis of GenoMic Annotation (MAGMA) in Functional Mapping and Annotation (FUMA). Then, genetic instruments for PhenoAgeAccl and adverse outcomes following surgery were identified to conduct a summary-based Mendelian Randomisation [Assumptions are listed in Figure 1]. Results GWAS for adverse outcomes following major surgery identified 3 lead SNPs (rs1360618, rs11848297, rs7978) [Figure 2]. Gene-set analysis showed enrichment in the immune system, cell migration, fear response, and iron metabolism. For PhenoAge Acceleration, gene-set analysis showed enrichment in immune system, and cell signalling pathways. 17 SNPs were selected which were used for the Two-Sample MR. We found no causal relationship, as the beta coefficient for Inverse variance weighted was -0.308 [standard error - 0.075, p=0.68]. For every 5-year increase in PhenoAge compared to chronological age, the risk of emergency re-hospitalisation or death increased by 21% (HR 1.16 – 1.3, p < 0.001). Conclusion Gene-set analysis for adverse outcomes following major surgery showed enrichment in immune system, iron metabolism, and fear response which have all recently been associated with mortality post-surgery in cohort studies(4,5). No causal relationship between PhenoAgeAccel and mortality post-surgery was established.Mendelian Randomisation assumptionsManhattan plot adverse-outcomes post-sur

LCoRL regulates growth and metabolism.

Genome Wide Association Studies (GWAS) in humans and livestock have identified genes associated with metabolic traits. However, the causality of many of these genes on metabolic homeostasis is largely unclear due to lack of detailed functional analyses. Here we report Ligand Dependent Corepressor-Like (LCoRL) as a metabolic regulator for body weight and glucose homeostasis. Although GWAS data show that LCoRL is strongly associated with body size, glucose homeostasis, and other metabolic traits in humans and livestock, functional investigations had not been performed. We generated Lcorl knockout mice (Lcorl-/-) and characterized the metabolic traits. We found that Lcorl-/- pups are born smaller than the wildtype littermates (WT) before reaching normal weight by 7-9 weeks of age. While aging, Lcorl-/- mice remain lean compared to WT mice, which is associated with a decrease in daily food intake. Glucose tolerance and insulin sensitivity are improved in Lcorl-/- mice. Mechanistically, this stunted growth is linked to a reduction of circulating levels of insulin like growth factor-1. The expression of the genes downstream of growth hormone signaling and the genes involved in glucose and lipid metabolism are altered in the liver of Lcorl-/- mice. Furthermore, Lcorl-/- mice are protected against a high-fat diet (HFD) challenge and show reduced exercise capacity in an exercise stress test. Collectively, our results are congruent with many of the metabolic parameters linked to the Lcorl locus as reported in GWAS in humans and livestock.

Population-aware permutation-based significance thresholds for genome-wide association studies

Abstract Motivation Permutation-based significance thresholds have been shown to be a robust alternative to classical Bonferroni significance thresholds in genome-wide association studies for skewed phenotype distributions. The recently published method permGWAS introduced a batch-wise approach to efficiently compute permutation-based genome-wide association studies. However, running multiple univariate tests in parallel leads to many repetitive computations and increased computational resources. More importantly, traditional permutation methods that permute only the phenotype break the underlying population structure. Results We propose permGWAS2, an improved method that does not break the population structure during permutations and uses an elegant block matrix decomposition to optimize computations, thereby reducing redundancies. We show on synthetic data that this improved approach yields a lower false discovery rate for skewed phenotype distributions compared to the previous version and the commonly used Bonferroni correction. In addition, we re-analyze a dataset covering phenotypic variation in 86 traits in a population of 615 wild sunflowers (Helianthus annuus L.). This led to the identification of dozens of novel associations with putatively adaptive traits, and removed several likely false-positive associations with limited biological support. Availability permGWAS2 is open-source and publicly available on GitHub for download: https://github.com/grimmlab/permGWAS. Supplementary information Supplementary data are available at Bioinformatics Advances online.

Functional genomics implicates natural killer cells in the pathogenesis of ankylosing spondylitis

Multiple lines of evidence indicate that ankylosing spondylitis (AS) is a lymphocyte-driven disease. However, which lymphocyte populations are critical in AS pathogenesis is not known. In this study, we aimed to identify the key cell types mediating the genetic risk in AS using an unbiased functional genomics approach. We integrated genome-wide association study (GWAS) data with epigenomic and transcriptomic datasets of human immune cells. To quantify enrichment of cell type-specific open chromatin or gene expression in AS risk loci, we used three published methods - LDSC-SEG, SNPsea, scDRS - that have successfully identified relevant cell types in other diseases. Natural killer (NK) cell-specific open chromatin regions are significantly enriched in heritability for AS, compared to other immune cell types such as T cells, B cells, and monocytes. This finding was consistent between two AS GWAS. Using RNA-seq data, we validated that genes in AS risk loci are enriched in NK cell-specific gene expression. Using the human Space-Time Gut Cell Atlas, we also found significant upregulation of AS-associated genes predominantly in NK cells. We performed co-localization analyses between GWAS risk loci and genetic variants associated with gene expression (eQTL) to find putative target genes. This revealed four AS risk loci affecting regulation of candidate target genes in NK cells: two known loci, ERAP1 and TNFRSF1A, and two under-studied loci, ENTR1 (aka SDCCAG3) and B3GNT2. Our findings suggest that NK cells may play a crucial role in AS development and highlight four putative target genes for functional follow-up in NK cells.

Unsupervised electrocardiogram feature extraction using deep learning empowers discovery of genetic and phenotypic determinants of cardiac electrophysiology

Abstract Background Conventional analysis of the electrocardiogram (ECG) is based on the extraction of human-defined, visually recognisable features. However, these are not optimised to capture all the information content in the ECG. The variational autoencoder (VAE), a form of unsupervised machine learning, can address this shortcoming by computationally extracting comprehensive and interpretable new ECG features, called latent factors. These latent factors provide a low-dimensional representation of the ECG that maximises capture of data content. This approach could expand our understanding of electrophysiology by identifying novel genetic and phenotypic traits associations with the ECG. Purpose This study aims to uncover genetic determinants of VAE latent factors, comparing them to determinants of conventional, human-derived ECG parameters to gain novel insights into cardiac electrophysiology and related diseases. Methods Over one million median beat ECGs derived from a United States (US) based secondary care centre were used to train a machine learning VAE model, with external validation in the UK Biobank (UKB). We performed common and rare variant association studies for VAE latent factors and conventional ECG traits on quality-controlled UKB data. Associated genetic variants were compared to loci for conventional ECG parameters available in the UKB and literature. Novel GWAS associations were validated in a withheld subset of the UKB cohort. Additionally, we compared the associations of the VAE latent factors and conventional ECG traits with phenotypic traits, disease codes and echocardiographic traits. Results Utilising median beat ECGs, the VAE model identified 20 features, with partial correlations to traditional ECG traits. A genome-wide association study (GWAS) identified 105 associated genomic regions, compared with 62 regions identified with conventional ECG traits on the same dataset. Most discovered loci have been previously associated with ECG features in larger GWAS, providing a positive control validation. We also discovered and validated five genomic regions not previously linked to the ECG (mapped to TRIOBP, EFEMP1, NEDD9, ATP10A and P2RX1). We found rare variant associations for seven genes (NEK6, IL17RA, NME7, MYBPC3, CCT8, ADAMTS6 and SCN5A). The latent factors uncovered associations with 158 phenotypic traits (2093 phenotypes tested) and 147 disease codes (2074 disease phecodes tested) which were not identified by ECG traits. Out of 38 echocardiographic traits, 35 showed higher correlations with the latent factors, while six exhibited significant associations exclusively with the latent factors. Conclusion Our study demonstrates that VAE-derived latent factors outperform conventional ECG parameters in uncovering novel genetic and phenotypic associations. Our findings underscore the opportunity for optimised data-driven feature extraction to enhance our understanding of cardiac electrophysiology.

The genetic architecture of exercise-induced ventricular ectopy

Abstract Background Previously, we have shown that the frequency of exercise-induced premature ventricular contractions (PVCs) during and after exercise are associated with increased risk of major adverse cardiovascular events in asymptomatic individuals from the UK Biobank study1. The underlying mechanisms are, however, unclear. Several electrocardiographic markers associated with arrhythmogenesis have a known hereditary component, but the genetic underpinnings of exercise-induced ventricular ectopy have not been investigated. Purpose Our aim was to explore the genetic basis of premature ventricular ectopy during and after exercise and unveil plausible candidate genes and biological mechanisms. Methods We conducted a genome-wide association study (GWAS) for PVC frequency during exercise and recovery in >44,000 asymptomatic individuals of European ancestry without known cardiovascular disease from the UK Biobank study. As the distribution of PVC frequency is highly skewed (e.g. absence of negative values and excess of zero values), we first applied Gaussian transformation. Simulation studies were performed to justify that this approach would lead to valid inference. In the GWAS model, we included sex, age, UK Biobank genotyping array, and the first 10 genetic principal components as covariates. Statistical power was boosted by joint analysis of exercise and recovery PVC traits using multi-trait analysis of GWAS (MTAG)2. Genetic risk scores were constructed in independent samples from UK Biobank (N=341,948) and were tested for association with heart failure and life-threatening ventricular arrhythmias. Results We identified 4 loci for PVC frequency during exercise, and 1 locus for PVC frequency during recovery (Picture 1). Candidate genes included CRIM1, FLNC, and BAG3 for which mutations have been linked in the pathogenesis of (arrhythmogenic) cardiomyopathy. Subjects in the top 10% of the genetic risk scores had a significantly higher risk of heart failure and life-threatening ventricular arrhythmias compared to the bottom 90%, (odds ratio (OR) (95% confidence interval): 1.14 (1.08 – 1.20), p < 0.001; and 1.13 (1.02 – 1.24), p < 0.001, respectively). Conclusions Ventricular ectopic burdens during exercise and recovery have a significant heritable component. Candidate genes highlight a possible shared genetic background with inherited cardiomyopathy and a genetic risk score was associated with heart failure and ventricular arrhythmias. These findings substantial advance our understanding of the genetic basis of exercise-induced ventricular ectopy in a population-based cohort without cardiovascular disease.Miami plot of ventricular ectopic burden

Potential use of pharmacogenetics in cardiology: genome-wide association studies of amiodarone-induced thyroid disorders

Abstract Background Amiodarone is a commonly prescribed antiarrhythmic drug used to manage supraventricular and ventricular arrhythmias. The use of amiodarone is associated with a broad range of adverse effects, including amiodarone-induced hypothyroidism (AIH) or amiodarone-induced thyrotoxicosis (AIT) (1). Both conditions are linked with increased mortality and have no known reliable risk predictors (2). Purpose We investigated the genetic underpinnings of amiodarone-induced thyroid disorders, and the clinical validity and utility of screening for genetic variants. Methods We conducted the first genome-wide association studies (GWAS) of AIH and AIT using data from three large biobanks. We compared the AUCs for SNPs identified through the above GWAS with polygenic risk scores for both hypo- and hyperthyroidism. Finally, we tested the clinical validity and utility of screening for risk variants. Results We found two risk loci for AIH. The lead variant at the first locus, rs1443438, was intronic to FOXE1 (OR 0.39, MAF = 35%, P = 1.59 × 10-42). The other lead variant, rs2209796, was located 93 kb upstream to FOXA2 (OR 0.56, MAF = 29%, P = 2.64 × 10-15: Figure 1). For AIT, we found one risk locus. The lead variant, rs2268803, was intronic to CAPZB (OR 1.81, MAF = 42%, P = 2.43 × 10-8: Figure 1). Next, we added GWAS SNPs or polygenic risk scores to risk prediction models consisting of age, sex and 4 principal components (PCs), and compared AUC estimates between risk models. Using GWAS SNPs, we found that the AIT variant increased the AUC with 6% (95%CI 1.1 – 10.3, AUC: 0.72) and 9% (95%CI 6.2 – 11.2, AUC: 0.75) for AIH, compared to the benchmark model. Both GWAS SNP models were superior to polygenic risk scores in terms of AUC (P for difference <0.05). The positive and negative predictive value (PPV and NPV) for the AIT-variant was 14.6% and 95.7%, respectively. For the AIH-variants combined, we found a PPV of 17% and a NPV of 95%. Conclusions Common genetic variants are associated with both amiodarone-induced hypo- and hyperthyroidism and are at this point, superior to polygenic risk scores in predicting amiodarone-induced thyroid disorders. These findings add to the potentials of pharmacogenetics in cardiology. Figure 1. Double Manhattan plot of amiodarone induced thyroid disorders. The Y-axis is the negative log10 of the P-value for each single nucleotide polymorphism. The X-axis is the chromosome, with positions on each chromosome ranging from lower to higher (left to right). The dotted line marks the threshold for genome-wide significance. Amiodoarone-induced hypothyroidism at the top, and amiodarone-induced thyrotoxicosis at the bottom.Figure 1

Micro-capture-C in vessel wall cell types identifies mechanisms underlying coronary artery disease susceptibility loci

Abstract Genetics is estimated to account for 50% of Coronary Artery Disease (CAD) risk. Genome-Wide Association Studies (GWAS) have revealed 100s of genomic loci which are associated with CAD (1). However, moving from GWAS signals to causal genes has proved challenging, slowing progress in developing interventions for CAD and other diseases. The greatest challenge in interpreting GWAS data is that only 10% of variants are in protein-coding regions and 90% are in "non-coding" genomic regions. Many of these non-coding variants are found in regulatory elements such as enhancers. Enhancers are distal regulatory elements that bind transcription factors and regulate genes up to 1 million base pairs away. This is achieved by DNA looping bringing enhancers into physical proximity with target genes. However, it is not possible to predict which gene(s) a given enhancer will regulate based on genomic location alone. We aimed to develop a generalisable workflow to solve CAD GWAS signals accurately and sensitively, by identifying causal enhancer SNPs and linking them mechanistically to output genes in a cell-type specific manner. We combined cell-specific epigenomic mapping with machine learning to prioritise 21,959 SNPs from 861 candidate CAD loci, in endothelial cells (HUVECs) and smooth muscle cells (HCASMCs). Micro-Capture-C (MCC) was used to resolve enhancer-promoter interactions at base-pair resolution (2). MNase footprinting determined differential transcription factor binding at specific enhancer SNPs. In HCASMCs, 20 putative enhancers containing CAD SNPs were found to interact with 32 genes, and in HUVECs 25 enhancers interacted with 48 genes. These genes were enriched for pathways including TGFb signalling, RUNX3 activity, and MAPK6/MAPK4 signalling. To prove causality at specific loci, heterozygous SNPs were identified in the individual primary cell lines. The interaction frequency at these elements was statistically compared between risk and non-risk alleles. For example, at the 9p21.3 locus, an enhancer containing a non-risk T allele was found to interact more strongly with the CDKN2B gene, compared to the risk G allele. Interestingly, this imbalance between allele interaction was potentiated by stimulation with TGFb, a key mediator in atherosclerosis. Furthermore, the transcription factor footprint at the SNP location was altered by the risk allele, indicating the mechanistic cause of the CAD-associated change in CDKN2B expression. We demonstrate a combined computational and experimental approach to identify causal non-coding variants in CAD risk loci, at scale. Among the key outputs of this study is the identification of the risk variant (rs1537373), output gene (CDKN2B), and specific cell type (HCASMC) underlying the 9p21.3 locus association with CAD risk. Using this approach to translate GWAS signals into cell-type specific outputs genes, will lead to a better understanding of the pathogenesis of CAD, and new therapeutic opportunities.

A proteome-wide association study identifies putative causal proteins for breast cancer risk.

Genome-wide association studies (GWAS) have identified more than 200 breast cancer risk-associated genetic loci, yet the causal genes and biological mechanisms for most loci remain elusive. Proteins, as final gene products, are pivotal in cellular function. In this study, we conducted a proteome-wide association study (PWAS) to identify proteins in breast tissue related to breast cancer risk. We profiled the proteome in fresh frozen breast tissue samples from 120 cancer-free European-ancestry women from the Susan G. Komen Tissue Bank (KTB). Protein expression levels were log2-transformed then normalized via quantile and inverse-rank transformations. GWAS data were also generated for these 120 samples. These data were used to build statistical models to predict protein expression levels via cis-genetic variants using the elastic net method. The prediction models were then applied to the GWAS summary statistics data of 133,384 breast cancer cases and 113,789 controls to assess the associations of genetically predicted protein expression levels with breast cancer risk overall and its subtypes using the S-PrediXcan method. A total of 6388 proteins were detected in the normal breast tissue samples from 120 women with a high detection false discovery rate (FDR) p value < 0.01. Among the 5820 proteins detected in more than 80% of participants, prediction models were successfully built for 2060 proteins with R > 0.1 and P < 0.05. Among these 2060 proteins, five proteins were significantly associated with overall breast cancer risk at an FDR p value < 0.1. Among these five proteins, the corresponding genes for proteins COPG1, DCTN3, and DDX6 were located at least 1 Megabase away from the GWAS-identified breast cancer risk variants. COPG1 was associated with an increased risk of breast cancer with a p value of 8.54 × 10-4. Both DCTN3 and DDX6 were associated with a decreased risk of breast cancer with p values of 1.01 × 10-3 and 3.25 × 10-4, respectively. The corresponding genes for the remaining two proteins, LSP1 and DNAJA3, were located in previously GWAS-identified breast cancer risk loci. After adjusting for GWAS-identified risk variants, the association for DNAJA3 was still significant (p value of 9.15 × 10-5 and adjusted p value of 1.94 × 10-4). However, the significance for LSP1 became weaker with a p value of 0.62. Stratification analyses by breast cancer subtypes identified three proteins, SMARCC1, LSP1, and NCKAP1L, associated with luminal A, luminal B, and ER-positive breast cancer. NCKAP1L was located at least 1Mb away from the GWAS-identified breast cancer risk variants. After adjusting for GWAS-identified breast cancer risk variants, the association for protein LSP1 was still significant (adjusted p value of 6.43 × 10-3 for luminal B subtype). We conducted the first breast-tissue-based PWAS and identified seven proteins associated with breast cancer, including five proteins not previously implicated. These findings help improve our understanding of the underlying genetic mechanism of breast cancer development.

Utility of polygenic risk score for prediction of dilated cardiomyopathy and modulation of severity of LV dysfunction among cases

Abstract Background Dilated cardiomyopathy (DCM) is a predominant cause of heart failure and a leading indication for cardiac transplantation. While rare genetic variants are established as causative factors in familial forms of the disease, recent genome-wide association studies (GWAS) have also underscored the significant contribution of common genetic variation to DCM. Polygenic scores (PGS) constructed from GWAS have demonstrated promise in risk stratification for a range of common diseases and may similarly have value for risk prediction in DCM. However, data regarding the clinical utility of PGS in DCM remains sparse. Purpose We aimed to evaluate the utility of PGS for prediction of DCM and its association with left ventricular ejection fraction (LVEF) and family history. We further aimed to assess the contribution of PGS in DCM patients with established causative rare variants (genotype-positive) and those without (genotype-negative). Methods We used a previously-constructed PGS, which was developed from a GWAS meta-analysis (8387 DCM cases and 939161 controls) and multi-trait analysis (MTAG) with cardiac MRI traits. In the present analysis, we calculated PGS for a newly assembled cohort of 978 DCM cases from a major university medical centre and 7207 controls from a national register. To evaluate the clinical utility of PGS, we built logistic regression models and assessed the association between PGS and DCM status. Subsequently, we performed subgroup analyses, including: i) individuals of European ancestry, ii) non-European ancestry, iii) males, iv) females, v) genotype-positive participants(n=193), and vi) genotype-negative participants (n=294). We then used linear and logistic regression models, respectively, to assess the associations of PGS with LVEF and documented family history of DCM. Results PGS demonstrated a significant enrichment among DCM cases compared to controls (OR per SD 1.93, P = 9.47E-68). This association persisted with consistent effect estimates - ranging from OR 1.5 to 2.2 - across all examined subgroups and demonstrated sufficient statistical significance (all P&amp;lt;1.58E-06) (Figure 1). Genotype-negative cases had a significantly higher PGS compared to genotype-positive individuals (P = 0.0015), although PGS was significantly enriched in both groups as compared to controls. Furthermore, among DCM cases, higher PGS was significantly associated with lower LVEF at first presentation (P=0.03, beta=-0.5% per SD) (Figure 2). There was no significant association found between PGS and family history. Conclusions PGS is strongly associated with risk of DCM, and may modulate the severity of LV dysfunction among DCM cases. Polygenic burden contributes to DCM risk in both genotype-positive and negative cases, although the contribution is stronger in patients without known causative rare variants. Overall, our results suggest a potential clinical applicability for PGS in DCM risk stratification.Figure 1.Association of PGS with DCMFigure 2.Association of PGS with LVEF