Sparse dimensionality reduction approaches in Mendelian randomisation with highly correlated exposures

Exploring the impact of dietary factors on intracranial aneurysm risk: insights from Mendelian randomization analysis

BACKGROUND

While existing research has established a link between dietary habits and the incidence of intracranial aneurysms, the application of Mendelian randomization to explore this association remains largely uncharted.

METHODS

In our study, we analyzed a wide array of dietary factors using data from the IEU Open GWAS project, which included meat varieties, vegetarian foods, cereal and the frequency of alcohol intake. We included pooled intracranial aneurysm GWAS data from a comprehensive dataset of 7,495 cases. In MR analysis, we employed multiple Mendelian randomization techniques such as MR-Egger, Inverse Variance Weighted methods and rigorously controlled the false discovery rates through the Bonferroni correction across 10 dietary exposures.

RESULTS

Our analysis identified a significant association between cooked vegetables (OR: 9.939; 95% CI: 2.066 ~ 47.822; p = 0.0042) and an elevated risk of intracranial aneurysms. Besides, the initial analysis suggested a statistically significant association between the dried fruit (OR: 0.385; 95%CI: 0.159 ~ 0.935; p = 0.0350), frequency of alcohol intake (OR: 1.419; 95% CI: 1.039 ~ 1.937; p = 0.0276) and the risk of intracranial aneurysms. However, this significance was not sustained after applying the Bonferroni correction for multiple testing, indicating a need for cautious interpretation despite the initially promising findings.

CONCLUSION

This study identified a clear causal link between cooked vegetable intake and an increased risk of intracranial aneurysm, while suggesting a potential connection between the frequency of alcohol intake and the elevated risk, although this association did not reach statistical significance after multiple testing corrections.

Leveraging multimodal neuroimaging and GWAS for identifying modality-level causal pathways to Alzheimer's disease

The UK Biobank study has produced thousands of brain imaging-driven phenotypes (IDPs) collected from more than 40,000 genotyped individuals so far, facilitating the investigation of genetic and imaging biomarkers for brain disorders. Motivated by efforts in genetics to integrate gene expression levels with genome-wide association studies (GWASs), recent methods in imaging genetics adopted an instrumental variable (IV) approach to identify causal IDPs for brain disorders. However, several methodological challenges arise with existing methods in achieving causality in imaging genetics, including horizontal pleiotropy and high dimensionality of candidate IVs. In this work, we propose testing the causality of each brain modality (i.e., structural, functional, and diffusion MRI) for each gene as a useful alternative, which offers an enhanced understanding of the roles of genetic variants and imaging features on behavior by controlling for the pleiotropic effects of IDPs from other imaging modalities. We demonstrate the utility of the proposed method by using Alzheimer's GWAS data from the UK Biobank and the International Genomics of Alzheimer's Project (IGAP) study. Our method is implemented using summary statistics, which is available on GitHub.

Bayesian Mendelian Randomization Analysis for Latent Exposures Leveraging GWAS Summary Statistics for Traits Co-Regulated by the Exposures

Mendelian Randomization analysis is a popular method to infer causal relationships between exposures and outcomes, utilizing data from genome-wide association studies (GWAS) to overcome limitations of observational research by treating genetic variants as instrumental variables. This study focuses on a specific problem setting, where causal signals may exist among a series of correlated traits, but the exposures of interest, such as biological functions or lower-dimensional latent factors that regulate the observable traits, are not directly observable. We propose a Bayesian Mendelian randomization analysis framework that allows joint analysis of the causal effects of multiple latent exposures on a disease outcome leveraging GWAS summary-level association statistics for traits co-regulated by the exposures. We conduct simulation studies to show the validity and superiority of the method in terms of type I error control and power due to a more flexible modeling framework and a more stable algorithm compared to an alternative approach and traditional single- and multi-exposure analysis approaches not specifically designed for the problem. We have also applied the method to reveal evidence of the causal effects of psychiatric factors, including compulsive, psychotic, neurodevelopmental, and internalizing factors, on neurodegenerative, autoimmune, digestive, and cardiometabolic diseases.

A novel multivariable Mendelian randomization framework to disentangle highly correlated exposures with application to metabolomics

Mendelian randomization (MR) utilizes genome-wide association study (GWAS) summary data to infer causal relationships between exposures and outcomes, offering a valuable tool for identifying disease risk factors. Multivariable MR (MVMR) estimates the direct effects of multiple exposures on an outcome. This study tackles the issue of highly correlated exposures commonly observed in metabolomic data, a situation where existing MVMR methods often face reduced statistical power due to multicollinearity. We propose a robust extension of the MVMR framework that leverages constrained maximum likelihood (cML) and employs a Bayesian approach for identifying independent clusters of exposure signals. Applying our method to the UK Biobank metabolomic data for the largest Alzheimer disease (AD) cohort through a two-sample MR approach, we identified two independent signal clusters for AD: glutamine and lipids, with posterior inclusion probabilities (PIPs) of 95.0% and 81.5%, respectively. Our findings corroborate the hypothesized roles of glutamate and lipids in AD, providing quantitative support for their potential involvement.

An augmented Mendelian randomization approach provides causality of brain imaging features on complex traits in a single biobank-scale dataset

Mendelian randomization (MR) is an effective approach for revealing causal risk factors that underpin complex traits and diseases. While MR has been more widely applied under two-sample settings, it is more promising to be used in one single large cohort given the rise of biobank-scale datasets that simultaneously contain genotype data, brain imaging data, and matched complex traits from the same individual. However, most existing multivariable MR methods have been developed for two-sample setting or a small number of exposures. In this study, we introduce a one-sample multivariable MR method based on partial least squares and Lasso regression (MR-PL). MR-PL is capable of considering the correlation among exposures (e.g., brain imaging features) when the number of exposures is extremely upscaled, while also correcting for winner's curse bias. We performed extensive and systematic simulations, and demonstrated the robustness and reliability of our method. Comprehensive simulations confirmed that MR-PL can generate more precise causal estimates with lower false positive rates than alternative approaches. Finally, we applied MR-PL to the datasets from UK Biobank to reveal the causal effects of 36 white matter tracts on 180 complex traits, and showed putative white matter tracts that are implicated in smoking, blood vascular function-related traits, and eating behaviors.