disjunct_pvalues: Test for infected/control/beads - a placebo effect?

In elsayed-lab/hpgltools: A pile of (hopefully) useful R functions

Test for infected/control/beads – a placebo effect?

Description

This was a function I copied out of Keith/Hector/Laura/Cecilia's paper in which they sought to discriminate the effect of inert beads on macrophages vs. the effect of parasites. The simpler way of expressing it is: take the worst p-value observed for the pair of contrasts, infected/uninfected and beads/uninfected.

Usage

disjunct_pvalues(contrast_fit, cellmeans_fit, conj_contrasts, disj_contrast)

Arguments

contrast_fit	Result of lmFit.
cellmeans_fit	Result of a cellmeans fit.
conj_contrasts	Result from the makeContrasts of the first set.
disj_contrast	Result of the makeContrasts of the second set.

Details

The goal is therefore to find responses different than beads The null hypothesis is (H0): (infected == uninfected) | (infected == beads) The alt hypothesis is (HA): (infected != uninfected) & (infected != beads)

elsayed-lab/hpgltools documentation built on May 9, 2024, 5:02 a.m.