ldsep (version 2.1.5)

Dprime: Get the standardized composite D'.

Description

This function will either standardize by the maximum covariance conditional on the marginal genotype distribution, or by the maximum covariance conditional on the marginal allele frequencies.

Usage

Dprime(qmat, type = c("allele", "geno"), constrain = FALSE)

Value

A vector of length 2. The first element is the estimated D'. The second element is the normalization used.

Arguments

qmat

The observed joint genotype distribution.

type

Should we condition on the marginal genotype distribution (type = "geno"), or should we condition on the allele frequency (type = "allele")?

constrain

A logical. This option is only applicable when type = "allele". Should return an value that is equal to D' under HWE (FALSE) or a value that is constrained to lie between -1 and 1 (TRUE)? Defaults to FALSE.

Author

David Gerard

Details

Note that when type = "allele" and constrain = FALSE, the resulting D' is constrained to fall between -K and K, where K is the ploidy of the species. However, under HWE, this measure is equal to haplotypic D'. Using constrain = TRUE will result in a measure that is constrained to lie between -1 and 1, but it will not equal haplotypic D' under HWE.

Using type = "geno" is its own thing and will not equal D' generally under HWE. When type = "geno", then the the constrain parameter has no effect.

Examples

K <- 6
qmat <- matrix(stats::runif((K+1)^2), nrow = K+1)
qmat <- qmat / sum(qmat)
Dprime(qmat, type = "geno")
Dprime(qmat, type = "allele")