Limma-trend is shown to compute approximate partially Bayes p-values, with new nonparametric maximum likelihood procedures that asymptotically control FDR under dense signals despite trend misspecification.
arXiv preprint arXiv:2209.13686 , year=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
How does limma-trend work? An empirical partially Bayes perspective
Limma-trend is shown to compute approximate partially Bayes p-values, with new nonparametric maximum likelihood procedures that asymptotically control FDR under dense signals despite trend misspecification.