postInteraction {ArrayTools} | R Documentation |
Based on the result from the interaction test by looking at the result from the regressResult object, this function partitions tne orignal data, an ExpressionSetinto groups, one contains the genes without the interaction and others contains the genes with the interaction across different level of covariates.
postInteraction(eSet, regressObject, mainVar, compare1, compare2, method = regressionMethod(regressObject), adj = adjustment(regressObject))
eSet |
an ExpressionSet |
regressObject |
a regressResult |
mainVar |
variable of main interest |
compare1 |
the first value of the mainVar . For example, suppose that
mainVar is "drug", and there are three unique values: "drug1",
"drug2", and "placebo". You would like to compare "drug1" to "drug2".
Then you would use "drug1" as compare1 |
compare2 |
Based on the example for compare1 , "drug2" will be
the compare2 |
method |
It is used to run regression within each level of the effect modifier. choose the follwoing three options: "limma" (LIMMA), "regression" (ordinary linear regression), "permutation" (permutation test) |
adj |
adjustment method for multiple comparison test, including "holm", "hochberg", "hommel", "bonferroni", "BH", "BY", "fdr", "none". The default value is "none". Type help(p.adjust) for more detail. |
an interactionResult
class. The first component contains all the result
for all the genes. The second component contains the genes without the interaction
effect. The rest of the components contains genes with the interactions.
Xiwei Wu, Arthur Li
data(eSetExample) design.int<- new("designMatrix", target=pData(eSetExample), covariates = c("Treatment", "Group"), intIndex = c(1, 2)) contrast.int<- new("contrastMatrix", design.matrix = design.int, interaction=TRUE) result.int<- regress(eSetExample, contrast.int) sigResult.int <- selectSigGene(result.int) intResult <- postInteraction(eSetExample, sigResult.int, mainVar ="Treatment", compare1 = "Treated", compare2 = "Control")