Upper bound for the number of false discoveries among most significant items

Usage

curveMaxFP(p.values, thr)

Arguments

p.values

A vector containing s p-values

thr

A vector of $K$ JER-controlling thresholds

Value

A vector of size $s$ giving an joint upper confidence bound on the number of false discoveries among the $k$ most significant items for all $k \in \{1\ldots s\}$.

Details

The time and space complexity of this function is O(s), which is optimal since s is the length of the returned vector.

References

Enjalbert-Courrech, N. & Neuvial, P. (2022). Powerful and interpretable control of false discoveries in two-group differential expression studies. Bioinformatics. doi: 10.1093/bioinformatcs/btac693

Author

Gilles Blanchard, Nicolas Enjalbert-Courrech, Pierre Neuvial and Etienne Roquain