R/multtest_adjust.R
dcAdjust.RdAdjust for multiple hypothesis testing after performing
statistical tests using dcTest. This can be performed using a
method provided by the users. p.adjust is used by default.
dcAdjust(dcpvals, f = stats::p.adjust, ...)a matrix, the result of the dcTest function. The
results should be passed as produced by the function and not modified in
intermediate steps
a function, the function to be used for adjustment. p.adjust
from the stats package is the default with the specific adjustment
method 'fdr' used. The range of available methods can be accessed using
p.adjust.methods. Custom functions should accept a numeric vector of
p-values as the first argument
additional parameters to the adjustment function such as
method
a matrix, of adjusted p-values (or scores in the case of DiffCoEx and EBcoexpress) representing significance of differential associations.
Ensure that the p-value matrix passed to this function is the one
produced by dcTest. Any modification to the result matrix will
result in failure of the function.
This method applies the adjustment method only to one triangle of the matrix to ensure adjustment is not performed for duplicated tests (symmetric matrix). As results from the DiffCoEx and EBcoexpress do not produce p-values, this method does not change anything thereby returning the original matrix.
x <- matrix(rnorm(60), 2, 30)
cond <- rep(1:2, 15)
zscores <- dcScore(x, cond)
pvals <- dcTest(zscores, emat = x, condition = cond)
dcAdjust(pvals, p.adjust, method = 'fdr')
#> 1 2
#> 1 NA 0.6520118
#> 2 0.6520118 NA
#> attr(,"cor.method")
#> [1] "pearson"
#> attr(,"call")
#> z.score(emat = emat, condition = condition)
#> attr(,"dc.method")
#> [1] "zscore"
#> attr(,"dc.test")
#> [1] "two tailed z-test (adj)"