Mendelian randomization with invalid instruments: effect ...

Mendelian randomization with invalid instruments: effect ...

Mendelian randomization with invalid instruments: Egger regression and Weighted Median Approaches David Evans What is the problem?

Mendelian Randomization (MR) uses genetic variants to test for causal relationships between phenotypic exposures and disease-related outcomes Due to the proliferation of GWAS, it is increasingly common for MR analyses to use large numbers of genetic variants

Increased power but greater potential for pleiotropy Pleiotropic variants affect biological pathways other than the exposure under investigation and therefore can lead to biased causal estimates and false positives under the null Two Sample MR:

Single Variants Wald = Beta-GY BetaGX Causal estimate using Wald method: Two Sample MR: Multiple Variants Causal estimate using IVW

from summarised data: (Approximates TSLS) MR with direct pleiotropy . Single variant Wald estimate: Multiple variant

TSLS / IVW : Egger Regression: Central concept In Mendelian Randomization when multiple genetic variants are being used as IVs, Egger regression can: o Identify the presence of directional pleiotropy (biasing the IV estimate) o provide a less biased causal estimate

(in the presence of pleiotropy) InSIDE Assumption Relaxing MRs assumptions . W Example:

SNP outcome association ALL INVALID INSTRUMENTS INSIDE ASSUMPTION SATISFIED SNP exposure association InSIDE:

SNP outcome association Bias of ratio estimator SNP exposure association Egger regression: Intercept not constrained to zero Eggers test assesses whether the intercept term is

significantly different from zero. The estimated values of the intercept can be interpreted as the average pleiotropic effect across all genetic variants. An intercept term different from zero indicates directional pleiotropy SNP-Height SNP-Lung Function Height and lung function

SNP - Height Causal estimate IVW = 0.59 (95% CI: 0.50, 0.67 ) Egger = 0.58 (95% CI: 0.50, 0.67); intercept -0.001 p=0.5 BP and Coronary Disease FUNNEL PLOTS

Visual evidence for asymmetry BP and Coronary Disease FUNNEL PLOTS BP and Coronary Disease Scatter Plots Egger test for intercept p=0.2

Egger test for intercept p=0.054 BP and Coronary Disease FUNNEL PLOTS IVW= 0.054 logOR/mmHg p=4x10-6 Egger =0.015 logOR/mmHg p=0.6 IVW= 0.083 logOR/mmHg p=1x10-5

Egger =-0.024 logOR/mmHg p=0.7 Simple Median Method Order instrumental variables estimates and take the median Like all subsequent estimators it enjoys a 50% breakdown limit Weighted Median Method

^2 j ___ wj = 2 Yj wj ____ wj =

jwj Penalized Weighted Median Method ough the invalid IVs do not contribute directly to the median estimate, they do uence it in small samples Like all subsequent estimators it enjoys a 50% breakdown limit Penalized Weighted

Median Method One way of minimizing this problem is down-weighting the contribution to the analysis of genetic variants with heterogeneous ratio estimates Heterogeneity between estimates can be quantified by Cochranes Q statistic: ^ ^2 Q = jQj =jwj( j - )

The Q statistic has a chi-squared distribution on J 1 degrees of freedom under the null hypothesis of no heterogeneity Each individual component of Q has a chi-square distribution with 1 df. Bowden proposes using a one sided upper P value (denoted qj): wj* = wj min(1, 20qj) Penalized Weighted Median Method

References Mendelian randomization with invalid instruments : effect estimation and bias detection through Egg er regression. Bowden J, Davey Smith G, Burgess S. Int J Epidemiol. 2015 Apr;44(2):512-25. Consistent Estimation in Mendelian Randomizatio n with Some Invalid Instruments Using a Weighte d Median Estimator. Bowden J, Davey Smith G, Haycock

PC, Burgess S. Genet Epidemiol. 2016 May;40(4):304-14.

Recently Viewed Presentations

  • Chordates - Teacher.co.ke

    Chordates - Teacher.co.ke

    Phylum Chordata. By the end of the Cambrian period, 540 million years ago, an astonishing variety of animals inhabited Earth's oceans. One of these types of animals gave rise to vertebrates, one of the most successful groups of animals.
  • Kudos AD - AGCAS

    Kudos AD - AGCAS

    Kudos AD. Personal development and career guidance tool for undergraduates and adults. Support users to define and develop their career objectives. Personal profile developed through: Interest inventory. Skills analysis. Supports development of self awareness through understanding of core interests and...
  • WCMS and Campus Web Search Updates

    WCMS and Campus Web Search Updates

    Liam Morland - Opening up some sites so all CAS users can authenticate. Chris Shantz- Prototypes for the IT Strategic Plan site and access control with Organic Groups. Kris Olafson - Feature requests site. #watitis2012. Six content types. Project .
  • Happy people  Good investments A case of Shared

    Happy people Good investments A case of Shared

    Presentation. My focus today will be on the Shared Care platform, which we have developed in collaboration with IBM because we believe that this platform is a very good investment that leads to happy people.
  • Cost Estimating - MIT

    Cost Estimating - MIT

    Cost Estimating Critical for: Determining whether to make an investment to provide a product for the consumer market Deciding if a company should quote on a product for sale to another company Purpose of Estimating Establish the bid price of...
  • DAVID GEIGER BAE  STRUCTURAL OPTION DULLES TOWN CENTER

    DAVID GEIGER BAE STRUCTURAL OPTION DULLES TOWN CENTER

    granby tower norfolk, virginia senior thesis presentation 14 april 2008 DAVID GEIGER BAE - STRUCTURAL OPTION DULLES TOWN CENTER BUILDING ONE DULLES, VIRGINIA SENIOR THESIS PRESENTATION 14 APRIL 2009
  • Chapter 6: Developing Product and Brand Strategy

    Chapter 6: Developing Product and Brand Strategy

    Product Mix Width: Number of lines offered. Product line Depth: Number of products in a line. Product Mix and Product Lines Line extensions: occur when a company introduces additional items in the same product category under the same brand name...
  • Light Search and Rescue Operations CERT Basic Training

    Light Search and Rescue Operations CERT Basic Training

    Light Search and Rescue Operations CERT Basic Training Unit 5