the {qgen}R Documentation

Transformation of causal genetic parameters to observational parameters

Description

For theoretical investigations we need a function that takes the chosen causal genetic parameters and transforms them into observational parameters. Additionally the form of the object needs to be adjusted to fit the form of an object of class paraDATA. This function handles the following causal variance components: additive; dominance, maternal, micro-environmental; and the following sources of observable variance components: sire, dam, and individuals.

Usage

the(partitioning = "all",
                chN = 1, # number of characters
                enN = 1, # number of environments
                fbN = 1, # number of fixed blocks
                rbN = 1, # number of random blocks
                siN = 100, # number sires (within each block)
                daN = 6, # number of dams (within each sire)
                idN = 3, # number of individuals (within each dam)
                randomblockCor = matrix(0.5, chN*enN, chN*enN),
                randomblockVar = rep(100, chN*enN),
                additiveCor = matrix(0.5, chN*enN, chN*enN),
                additiveVar = rep(100, chN*enN),
                dominanceCor = matrix(0.5, chN*enN, chN*enN),
                dominanceVar = rep(100, chN*enN),
                maternalCor = matrix(0.5, chN*enN, chN*enN),
                maternalVar = rep(100, chN*enN),
                environmentalCor = matrix(0, chN*enN, chN*enN),
                environmentalVar = rep(100, chN*enN),
                ch.names = paste("ch",1:chN,sep=""),
                en.names = paste("en",1:enN,sep=""),
                fb.names = paste("fb",1:fbN,sep=""),
                fixe = array(0, dim=c(enN, chN, fbN)),
                miss = array(0, dim=c(enN, chN, fbN)),
                file=TRUE,
                path="~/qgen/")

Arguments

partitioning character string: The method of variance partitioning used throughout the analysis
"all"
(default) resampling is based on REML solutions but both ANOVA estimators are provided additionally
"REML"
uses the package lme4 to calculate REML solutions
"ANOVA"
calculates traditional ANOVA estimators; can only be used for investigating exactly one character in one environment
"ANOVAuw"
calculates ANOVA estimators using unweighted sums of squares; can only be used for investigating exactly one character in one environment
chN number of characters
enN number of environments
fbN number of fixed blocks
rbN number of random blocks, a random factor crossed with sire
siN number of sires
daN number of dams within sires
idN number of individuals within dams
randomblockCor for the random block effects
additiveCor for the additive genetic effects
dominanceCor for the dominance genetic effects
maternalCor for the maternal effects
environmentalCor for the microenvironmental effects
randomblockVar for the random blocks
additiveVar for the additive genetic effects
dominanceVar for the dominance genetic effects
maternalVar for the maternal effects
environmentalVar for the microenvironmental effects
fixe array with the mean of every fixedblock–environment–character combinations (sequence: character in environments in fixedblock, ex. fb1: en1ch1, en1ch2...en2ch1...enxchx fb2: en1ch1,...)
miss vector with the proportion of randomly missing values per environment–character combination; if scalar all are assumed to be equal
en.names vector of character strings: names for the environments
ch.names vector of character strings: names for the traits
fb.names vector of character strings: names for the fixed blocks
file logical flag: Should the object be written to a file called "the.rda"?
path character vector: containing a single path name

Details

This function transforms the causal sources of variance (additive genetic effects, dominance genetic effects, common (maternal) environmental effects, and individual (special) environmental effects) into the three observable variance components from a nested full–sib, half–sib mating design (North-Carolina Design I). Assumptions: no sources of epistatic variance; see Lynch & Walsh 1997, p.572.

Value

a paraDATA object: with the individual slots

orig fully specified
supl fully specified
para fully specified
data empty
spec empty

Note

name: theoretical investigations on chosen biological parameters

References

Lynch, M. and Walsh, B. (1997) Genetics and analysis of quantitative traits. Sinauer.

See Also

emp, paraDATA

Examples

## create a paraDATA-object
## (with a full para section and an empty DATA section)
the(file=FALSE)

[Package qgen version 0.03-02 Index]