coloc.test.summary: Colocalisation testing using regression coefficients

Description Usage Arguments Details Value Author(s)


Colocalisation testing supplying only regression coefficients and their variance-covariants matrices


  coloc.test.summary(b1, b2, V1, V2, k = 1,
    plot.coeff = TRUE, plots.extra = NULL,
    bayes = !is.null(bayes.factor), n.approx = 1001, = 0.95, bayes.factor = NULL, bma = FALSE)



regression coefficients for trait 1


regression coefficients for trait 2


variance-covariance matrix for trait 1


variance-covariance matrix for trait 2


Theta has a Cauchy(0,k) prior. The default, k=1, is equivalent to a uniform (uninformative) prior. We have found varying k to have little effect on the results.


TRUE if you want to generate a plot showing the coefficients from the two regressions together with confidence regions.


parameter set to TRUE when coloc.test is called by coloc.bma. DO NOT SET THIS WHEN CALLING coloc.test DIRECTLY!


list with 2 named elements, x and y, equal length character vectors containing the names of the quantities to be plotted on the x and y axes.

x is generally a sequence of theta and eta, with y selected from post.theta, the posterior density of theta, chisq, the chi-square values of the test, and lhood, the likelihood function.


Logical, indicating whether to perform Bayesian inference for the coefficient of proportionality, eta. If bayes.factor is supplied, Bayes factors are additionally computed for the specificed values. This can add a little time as it requires numerical integration, so can be set to FALSE to save time in simulations, for example.


Calculate Bayes Factors to compare specific values of eta. bayes.factor should either a numeric vector, giving single value(s) of eta or a list of numeric vectors, each of length two and specifying ranges of eta which should be compared to each other. Thus, the vector or list needs to have length at least two.,n.approx denotes the required level of the credible interval for eta. This is calculated numerically by approximating the posterior distribution at n.approx distinct values.


Typically this should be called from coloc.test() or coloc.bma(), but is left as a public function, to use at your own risk, if you have some other way to define the SNPs under test.


an object of class coloc, colocBayes or colocBMA


Chris Wallace

Search within the coloc package
Search all R packages, documentation and source code

Questions? Problems? Suggestions? or email at

Please suggest features or report bugs with the GitHub issue tracker.

All documentation is copyright its authors; we didn't write any of that.