2-factorial limma-style t-test
The aim of this function is to provide convenient acces to two-factorial (linear) testing withing the framework of makeMAList including the emprical Bayes shrinkage.
The input data 'datMatr' which should already be organized as limma-type MAList, eg using using makeMAList.
Note: This function uses the Bioconductor package limma.
test2factLimma(
datMatr,
fac1,
fac2,
testSynerg = TRUE,
testOrientation = "=",
addResults = c("lfdr", "FDR", "Mval", "means"),
addGenes = NULL,
silent = FALSE,
callFrom = NULL
)datMatr |
matrix or data.frame with lines as indenpendent series of measures (eg different genes) |
fac1 |
(character or factor) vector describing grouping elements of each line of 'datMatr' for first factor, must be of same langth as fac2 |
fac2 |
(character or factor) vector describing grouping elements of each line of 'datMatr' for second factor, must be of same langth as fac1 |
testSynerg |
(logical) decide if factor-interactions (eg synergy) should be included to model |
testOrientation |
(character) default (or any non-recignized input) '=', otherwise either '>','gerater','sup','upper' or '<','inf','lower' |
addResults |
(character) vector defining which types of information should be included to output, may be 'lfdr','FDR' (for BY correction), 'Mval' (M values), 'means' (matrix with mean values for each group of replicates) |
addGenes |
(matrix or data.frame) additional information to add to output |
silent |
(logical) suppress messages |
callFrom |
(character) allow easier tracking of message(s) produced |
object of class "MArrayLM" (from limma)
makeMAList, single line testing lmFit and the eBayes-family of functions in package limma
## example for testing change of ratio for 4 sets (AA-DD) of pairs of data
set.seed(2017); t8 <- matrix(round(rnorm(160,10,0.4),2),ncol=8,
dimnames=list(letters[1:20],c("AA1","BB1","CC1","DD1","AA2","BB2","CC2","DD2")))
t8[3:6,1:2] <- t8[3:6,1:2]+3 # augment lines 3:6 (c-f) for AA1&BB1
t8[5:8,5:6] <- t8[5:8,5:6]+3 # augment lines 5:8 (e-h) for AA2&BB2 (c,d,g,h should be found)
## via MAobj
maOb8 <- makeMAList(t8,MAf=gl(2,4,labels=c("R","G")))
fit8b <- test2factLimma(maOb8,c(1,1,1,1),c(0,0,1,1),testS=FALSE) # same result as below (fit8e)
limma::topTable(fit8b,coef=1,n=5) # effect for c,d,g&h
## explicit (long) way via limma:
fit8 <- limma::lmFit(maOb8, design= model.matrix(~ 0+factor(c(1,1,2,2))))
fit8e <- limma::eBayes(fit8)
limma::topTable(fit8e,coef=1,n=5) # effect for c,d,g&hPlease choose more modern alternatives, such as Google Chrome or Mozilla Firefox.