Skip to contents

Bayesian implementation of the MR-Egger multivariate model with choice of prior distributions fitted using JAGS.

Source: R/mr_egger_rjags.R
mr_egger_rjags.Rd

Bayesian implementation of the MR-Egger multivariate model with choice of prior distributions fitted using JAGS.

Usage

mr_egger_rjags(
  object,
  prior = "default",
  betaprior = "",
  sigmaprior = "",
  n.chains = 3,
  n.burn = 1000,
  n.iter = 5000,
  seed = NULL,
  rho = 0.5,
  ...
)

Arguments

object

A data object of class mr_format.

prior

A character string for selecting the prior distributions;

  • "default" selects a non-informative set of priors;

  • "weak" selects weakly informative priors;

  • "pseudo" selects a pseudo-horseshoe prior on the causal effect;

  • "joint" selects a joint prior on the intercept and slope.

betaprior

A character string in JAGS syntax to allow a user defined prior for the causal effect.

sigmaprior

A character string in JAGS syntax to allow a user defined prior for the residual standard deviation.

n.chains

Numeric indicating the number of chains used in the MCMC estimation, the default is 3 chains.

n.burn

Numeric indicating the burn-in period of the Bayesian MCMC estimation. The default is 1000 samples.

n.iter

Numeric indicating the number of iterations in the Bayesian MCMC estimation. The default is 5000 iterations.

seed

Numeric indicating the random number seed. The default is the rjags default.

rho

Numeric indicating the correlation coefficient input into the joint prior distribution. The default value is 0.5.

...

Additional arguments passed through to rjags::jags.model().

Value

An object of class eggerjags containing the following components:

AvgPleio

The mean of the simulated pleiotropic effect

CausalEffect

The mean of the simulated causal effect

StandardError

Standard deviation of the simulated causal effect

sigma

The value of the residual standard deviation

CredibleInterval

The credible interval for the causal effect, which includes the lower (2.5%), median (50%) and upper intervals (97.5%)

samples

Output of the Bayesian MCMC samples

Priors

The specified priors

References

Bowden et. al., Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. International Journal of Epidemiology 2015. 44(2): p. 512-525. doi:10.1093/ije/dyv080

Examples

if (requireNamespace("rjags", quietly = TRUE)) {
fit <- mr_egger_rjags(bmi_insulin)
summary(fit)
plot(fit$samples)
# 90% credible interval
fitdf <- do.call(rbind.data.frame, fit$samples)
cri90 <- sapply(fitdf, quantile, probs = c(0.05, 0.95))
print(cri90)
}
#> Prior : 
#> 
#>  Pleiotropy ~ dnorm(0, 1E-3) 
#>  Estimate ~ dnorm(0, 1E-3) 
#>  sigma ~ dunif(.0001, 10) 
#> 
#> Estimation results: 
#>  
#>  MCMC iterations = 6000 
#>  Burn in = 1000 
#>  Sample size by chain = 5000 
#>  Number of Chains = 3 
#>  Number of SNPs = 14 
#>  
#> Inflating Parameter: 7.63072 
#> 
#>                  Estimate         SD       2.5%         50%      97.5%
#> Avg Pleio     -0.05423754 0.03587126 -0.1249326 -0.05289878 0.01408735
#> Causal Effect  3.74906123 2.12741936 -0.2994558  3.67181831 7.95763439

#>      Estimate   Pleiotropy    sigma
#> 5%  0.3637214 -0.114480488 5.686757
#> 95% 7.3196728  0.003112578 9.600718