HinksEtAl2010 {bayesmeta} | R Documentation |
Log odds ratios indicating association of a genetic variant (CCR5) with juvenile idiopathic arthritis (JIA).
data("HinksEtAl2010")
The data frame contains the following columns:
study | character | publication identifier |
year | numeric | publication year |
country | character | country |
or | numeric | odds ratio (OR) |
or.lower | numeric | lower 95 percent confidence bound for OR |
or.upper | numeric | upper 95 percent confidence bound for OR |
log.or | numeric | logarithmic OR |
log.or.se | numeric | standard error of logarithmic OR |
Results from a genetic association study (Hinks et al, 2010) were combined with data from two additional studies (Prahalad et al., 2006; Lindner et al., 2007) in order to determine the combined evidence regarding the association of a particular genetic marker (CCR5) with juvenile idiopathic arthritis (JIA).
A. Hinks et al. Association of the CCR5 gene with juvenile idiopathic arthritis. Genes and Immunity, 11(7):584-589, 2010. doi: 10.1038/gene.2010.25.
S. Prahalad et al. Association of two functional polymorphisms in the CCR5 gene with juvenile rheumatoid arthritis. Genes and Immunity, 7:468-475, 2006. doi: 10.1038/sj.gene.6364317.
E. Lindner et al. Lack of association between the chemokine receptor 5 polymorphism CCR5delta32 in rheumatoid arthritis and juvenile idiopathic arthritis. BMC Medical Genetics, 8:33, 2007. doi: 10.1186/1471-2350-8-33.
C. Roever, G. Knapp, T. Friede. Hartung-Knapp-Sidik-Jonkman approach and its modification for random-effects meta-analysis with few studies. BMC Medical Research Methodology, 15:99, 2015. doi: 10.1186/s12874-015-0091-1.
data("HinksEtAl2010") ## Not run: # perform meta analysis based on weakly informative half-normal prior: bma01 <- bayesmeta(y = HinksEtAl2010$log.or, sigma = HinksEtAl2010$log.or.se, labels = HinksEtAl2010$study, tau.prior = function(t){dhalfnormal(t,scale=1.0)}) # perform meta analysis based on slightly more informative half-normal prior: bma02 <- bayesmeta(y = HinksEtAl2010$log.or, sigma = HinksEtAl2010$log.or.se, labels = HinksEtAl2010$study, tau.prior = function(t){dhalfnormal(t,scale=0.5)}) # show heterogeneity posteriors: par(mfrow=c(2,1)) plot(bma01, which=4, prior=TRUE, taulim=c(0,1)) plot(bma02, which=4, prior=TRUE, taulim=c(0,1)) par(mfrow=c(1,1)) # show heterogeneity estimates: rbind("half-normal(1.0)"=bma01$summary[,"tau"], "half-normal(0.5)"=bma02$summary[,"tau"]) # show q-profile confidence interval for tau in comparison: require("metafor") ma03 <- rma.uni(yi=log.or, sei=log.or.se, slab=study, data=HinksEtAl2010) confint(ma03)$random["tau",c("ci.lb","ci.ub")] # show I2 values in the relevant range: tau <- seq(0, 0.7, by=0.1) cbind("tau"=tau, "I2" =bma01$I2(tau=tau)) # show effect estimates: round(rbind("half-normal(1.0)" = bma01$summary[,"mu"], "half-normal(0.5)" = bma02$summary[,"mu"]), 5) # show forest plot: forestplot(bma02) # show shrinkage estimates: bma02$theta ## End(Not run)