calculateThreshold {iBMQ} | R Documentation |
In the context of multiple testing and discoveries, a popular approach is to use a common threshold leading to a desired false discovery rate (FDR). In the Bayesian paradigm, derivation of the PPA threshold is trivial and can be calculated using a direct posterior probability calculation as described in Newton et al. (2004).
calculateThreshold(prob, threshold)
prob |
matrix or data frame that contains Posterior Probability of Association (output of eqtlMcmc function). |
threshold |
The desired false discovery rate. |
cutoff |
The significance threshold value |
Newton, MA., Noueiry, A., Sarkar, D. and Ahlquist, P. (2004): "Detecting differential gene expression with a semiparametric hierarchical mixture method."Biometrics, 5(2), 155-176
data(PPA.liver) cutoff.liver <- calculateThreshold(PPA.liver, 0.2)