A robust and efficient method for Mendelian randomization with hundreds of genetic variants

Burgess, Stephen; Foley, Christopher N.; Allara, Elias; Staley, James R; Howson, Joanna M.M.

doi:10.1038/s41467-019-14156-4

Cited by 452 publications

(403 citation statements)

References 57 publications

Supporting

Mentioning

397

Contrasting

Order By: Relevance

“…The contamination mixture method assumes that only some of the genetic variants are valid IVs (Burgess, Foley, Allara, Staley, & Howson, 2020). We construct a likelihood function from the ratio estimates.…”

Section: Methodsmentioning

confidence: 99%

A comparison of robust Mendelian randomization methods using summary data

Slob

Burgess

2020

Genetic Epidemiology

Self Cite

427

300

View full text Add to dashboard Cite

The number of Mendelian randomization (MR) analyses including large numbers of genetic variants is rapidly increasing. This is due to the proliferation of genome-wide association studies, and the desire to obtain more precise estimates of causal effects. Since it is unlikely that all genetic variants will be valid instrumental variables, several robust methods have been proposed. We compare nine robust methods for MR based on summary data that can be implemented using standard statistical software. Methods were compared in three ways: by reviewing their theoretical properties, in an extensive simulation study, and in an empirical example. In the simulation study, the best method, judged by mean squared error was the contamination mixture method. This method had well-controlled Type 1 error rates with up to 50% invalid instruments across a range of scenarios. Other methods performed well according to different metrics. Outlier-robust methods had the narrowest confidence intervals in the empirical example. With isolated exceptions, all methods performed badly when over 50% of the variants were invalid instruments. Our recommendation for investigators is to perform a variety of robust methods that operate in different ways and rely on different assumptions for valid inferences to assess the reliability of MR analyses. K E Y W O R D S causal inference, Mendelian randomization, pleiotropy, robust estimation, summary statistics ---This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

show abstract

Section: Methodsmentioning

confidence: 99%

A comparison of robust Mendelian randomization methods using summary data

Slob

Burgess

2020

Genetic Epidemiology

Self Cite

427

300

View full text Add to dashboard Cite

show abstract

“…As sensitivity analyses, we used the weighted median, MR-Egger, MR Pleiotropy RESidual Sum and Outlier (MR-PRESSO) and contamination mixture methods. 23,[27][28][29] In an additional sensitivity analysis, we excluded selfreported cancer in UK Biobank. We further conducted a sensitivity analysis using the SNP in the IGF1 gene (rs11111274), which was strongly associated with IGF-1 levels (7.59 × 10 −175 ), as genetic instrument.…”

Section: Introductionmentioning

confidence: 99%

Insulin‐like growth factor‐1 and site‐specific cancers: A Mendelian randomization study

et al. 2020

Self Cite

View full text Add to dashboard Cite

Insulin‐like growth factor‐1 (IGF‐1) is involved in several processes relevant to carcinogenesis. We used 416 single‐nucleotide polymorphisms robustly associated with serum IGF‐1 levels to assess the potential causal associations between this hormone and site‐specific cancers through Mendelian randomization. Summary‐level genetic association estimates for prostate, breast, ovarian, and lung cancer were obtained from large‐scale consortia including individuals of European‐descent. Furthermore, we estimated genetic associations with 14 site‐specific cancers in European‐descent individuals in UK Biobank. Supplementary analyses were conducted for six site‐specific cancers using summary‐level data from the BioBank Japan Project. Genetically predicted serum IGF‐1 levels were associated with colorectal cancer. The odds ratio (OR) per standard deviation increase of IGF‐1 levels was 1.11 (95% confidence interval [CI] 1.01‐1.22; P = .03) in UK Biobank and 1.22 (95% CI 1.09‐1.36; P = 3.9 × 10−4) in the BioBank Japan Project. For prostate cancer, the corresponding OR was 1.10 (95% CI 1.01‐1.21; P = .04) in UK Biobank, 1.03 (95% CI 0.97‐1.09; P = .41) in the prostate cancer consortium, and 1.08 (95% CI 0.95‐1.22; P = .24) in the BioBank Japan Project. For breast cancer, the corresponding OR was 0.99 (95% CI 0.92‐1.07; P = .85) in UK Biobank and 1.08 (95% CI 1.02‐1.13; P = 4.4 × 10−3) in the Breast Cancer Association Consortium. There was no statistically significant association between genetically predicted IGF‐1 levels and 14 other cancers. This study found some support for a causal association between elevated serum IGF‐1 levels and increased risk of colorectal cancer. There was inconclusive or no evidence of a causal association of IGF‐1 levels with prostate, breast, and other cancers.

show abstract

“…If data on genetic associations with multiple traits were used to cluster variants, then the clusters might be more precisely defined, but it would not be possible to determine which traits were driving the division into clusters without further analysis. We have previously demonstrated that a group of variants having similar causal estimates for the effect of HDL-cholesterol on CAD risk also had a distinct pattern of associations with blood cell traits, although without using a formal clustering method [30]. The associations with blood cell traits suggested a causal pathway relating to platelet aggregation.…”

Section: Discussionmentioning

confidence: 99%

MR-Clust: Clustering of genetic variants in Mendelian randomization with similar causal estimates

Kirk

Burgess

2019

Preprint

Self Cite

View full text Add to dashboard Cite

Motivation: Mendelian randomization is an epidemiological technique that uses genetic variants as instrumental variables to estimate the causal effect of a risk factor on an outcome.We consider a scenario in which causal estimates based on each variant in turn differ more strongly than expected by chance alone, but the variants can be divided into distinct clusters, such that all variants in the cluster have similar causal estimates. This scenario is likely to occur when there are several distinct causal mechanisms by which a risk factor influences an outcome with different magnitudes of causal effect. We have developed an algorithm MR-Clust that finds such clusters of variants, and so can identify variants that reflect distinct causal mechanisms. Two features of our clustering algorithm are that it accounts for uncertainty in the causal estimates, and it includes 'null' and 'junk' clusters, to provide protection against the detection of spurious clusters.Results: Our algorithm correctly detected the number of clusters in a simulation analysis, outperforming the popular Mclust method. In an applied example considering the effect of blood pressure on coronary artery disease risk, the method detected four clusters of genetic variants. A hypothesis-free search suggested that variants in the cluster with a negative effect of blood pressure on coronary artery disease risk were more strongly related to trunk fat percentage and other adiposity measures than variants not in this cluster.Availability and Implementation: MR-Clust can be downloaded from https://

show abstract

A robust and efficient method for Mendelian randomization with hundreds of genetic variants

Cited by 452 publications

References 57 publications

A comparison of robust Mendelian randomization methods using summary data

A comparison of robust Mendelian randomization methods using summary data

Insulin‐like growth factor‐1 and site‐specific cancers: A Mendelian randomization study

MR-Clust: Clustering of genetic variants in Mendelian randomization with similar causal estimates

Contact Info

Product

Resources

About