findFClog {factDesign}R Documentation

A function to find the fold change between two experimental conditions in a factorial experiment based on the linear model parameter estimates for the data that has been log-transformed.

Description

'findFClog' contructs a point estimate of fold change using the linear model coefficients in an lm object for data that has been log-transformed.

Usage

findFClog(model, lambdaNum, lambdaDenom)

Arguments

model An lm object.
lambdaNum A numeric vector of coefficients for the parameters to be used in the numerator of the fold change estimate.
lambdaDenom A numeric vector of coefficients for the parameters to be used in the denominator of the fold change estimate.

Value

A point estimate of the fold change between the experimental conditions specified in the lambdaNum and lambdaDenom vectors for data that has been log-transformed for the linear model estimation.

Author(s)

Denise Scholtens

See Also

par2lambda, findFC

Examples

data(estrogen)
ES <- pData(estrogen)[["ES"]]
TIME <- pData(estrogen)[["TIME"]]
fit <- lm(log(exprs(estrogen)[6,]) ~ ES + TIME + ES*TIME)
betaNames <- names(fit[["coef"]])
betas <- list(c("(Intercept)","ESP","TIME48h","ESP:TIME48h"),
                                        c("(Intercept)","ESP"))
coefs <- list(c(1,1),c(1,1))
lambda <- par2lambda(betaNames,betas,coefs)
findFClog(fit,lambda[1,],lambda[2,])




[Package Contents]