First, estimate empirical p-values based on a comparison of the actual and permuted test statistics. Next, estimate the proportion of true null hypotheses using the qvalue package as well as qvalues from the empirical p-values, using this value. If the estimated pi0 <= 0, then sequentially recalculates using increasingly conservative set of lambda values, until lambda = 0.5.
permQValue(dcObject, permObject, secondMat, testSlot, verbose = FALSE,
plotFdr = FALSE)A list containing a vectof of empirical p-values and a vector of q-values, both of the same length as the original actual test statistics.
The original S4 class object containing the test statistics to be extracted.
The array of matrices containing the null test statistics.
Logical, indicating whether a second matrix was used in the construction of this dcObject and permObject. If FALSE, the upper.tri of both are extracted to avoid double counting test statistics.
The slot of the dcObject to be removed for use as the actual test statistic.
Whether summaries of the q-value operations should be reported.
Allows for plotting of fdrtool p-value adjustment result OR empirical FDR q-value adjustment technique, if either of these are chosen. Requires fdrtool package OR qvalue package. Default = FALSE.