convertStartmeans {fitPoly}

A function to convert a set of mixture means from one ploidy to another

Description

convertStartmeans takes a set of means at one ploidy level (e.g. the fitted means for a tetraploid data set) and uses them to generate a set of means for another ploidy level (e.g. as startmeans for fitting triploid data for the same markers).

Usage

convertStartmeans(ploidy, origmeans)

Arguments

ploidy

The ploidy to which the means must be converted.

origmeans

A data.frame with a first column MarkerName, followed by <oldploidy+1> columns (names are ignored) that contain the ratio means for dosages 0 to <oldploidy>. Column MarkerName may not contain missing values. On each row the other columns must either all contain NA, or only non-NA values between 0 and 1 in strictly ascending order.

Details

The new means are calculated by linear interpolation between the old means on the asin(sqrt(x)) transformed scale and back-transformed to the original scale; the new means for dosage 0 are equal to the old, and the new means for dosage <ploidy> are equal to the old means for dosage <oldploidy>.

Value

A data.frame like origmeans with the same column MarkerName, now followed by <ploidy+1> columns with the new means.

Examples

# means from tetraploid data set:
tetrameans <- data.frame(MarkerName=c("mrk1", "mrk2"), mu0=c(0.02, 0.0),
mu1=c(0.2, 0.25), mu2=c(0.3, 0.5), mu3=c(0.4, 0.75), mu4=c(0.6, 1.0))
# convert to means for triploid data set:
trimeans <- convertStartmeans(ploidy=3, origmeans=tetrameans)
tetrameans
trimeans