postHocTest: Post-hoc tests for classic 1-way ANOVA

In samWieczorek/DAPAR: Tools for the Differential Analysis of Proteins Abundance with R

Post-hoc tests for classic 1-way ANOVA

Description

This function allows to compute a post-hoc test after a 1-way ANOVA analysis. It expects as input an object obtained with the function classic1wayAnova. The second parameter allows to choose between 2 different post-hoc tests: the Tukey Honest Significant Differences (specified as "TukeyHSD") and the Dunnett test (specified as "Dunnett").

Usage

postHocTest(aov_fits, post_hoc_test = "TukeyHSD")

Arguments

aov_fits	a list containing aov fitted model objects
post_hoc_test	a character string indicating which post-hoc test to use. Possible values are "TukeyHSD" or "Dunnett". See details for what to choose according to your experimental design.

Details

This is a function allowing to realise post-hoc tests for a set of proteins/peptides for which a classic 1-way anova has been performed with the function classic1wayAnova. Two types of tests are currently available: The Tukey HSD's test and the Dunnett's test. Default is Tukey's test. The Tukey HSD's test compares all possible pairs of means, and is based on a studentized range distribution. Here is used the TukeyHSD() function, which can be applied to balanced designs (same number of samples in each group), but also to midly unbalanced designs. The Dunnett's test compares a single control group to all other groups. Make sure the factor levels are properly ordered.

Value

a list of 2 dataframes: first one called "LogFC" contains all pairwise comparisons logFC values (one column for one comparison) for each analysed feature; The second one named "P_Value" contains the corresponding pvalues.

Author(s)

Hélène Borges

Examples

utils::data(Exp1_R25_prot, package='DAPARdata')
obj <- Exp1_R25_prot[1:1000]
level <- obj@experimentData@other$typeOfData
metacell.mask <- match.metacell(GetMetacell(obj), 'missing', level)
indices <- GetIndices_WholeMatrix(metacell.mask, op='>=', th=1)
obj <- MetaCellFiltering(obj, indices, cmd='delete')
anova_tests <- t(apply(Biobase::exprs(obj$new),1, classic1wayAnova, 
conditions=as.factor(Biobase::pData(obj$new)$Condition)))
names(anova_tests) <- rownames(Biobase::exprs(obj$new))
pht <- postHocTest(aov_fits = anova_tests)

samWieczorek/DAPAR documentation built on May 6, 2022, 5:30 p.m.