snp_scaleBinom: Binomial(n, p) scaling

In bigsnpr: Analysis of Massive SNP Arrays

Binomial(n, p) scaling

Description

Binomial(n, p) scaling where n is fixed and p is estimated.

Usage

snp_scaleAlpha(alpha = -1)

snp_scaleBinom(nploidy = 2)

Arguments

alpha	Assumes that the average contribution (e.g. heritability) of a SNP of frequency p is proportional to [2p(1-p)]^{1+\alpha}. The center is then 2 p and the scale is [2p(1-p)]^{-\alpha/2}. Default is -1.
nploidy	Number of trials, parameter of the binomial distribution. Default is 2, which corresponds to diploidy, such as for the human genome.

Details

You will probably not use this function as is but as the fun.scaling parameter of other functions of package bigstatsr.

Value

A new function that returns a data.frame of two vectors "center" and "scale" which are of the length of ind.col.

References

This scaling is widely used for SNP arrays. Patterson N, Price AL, Reich D (2006). Population Structure and Eigenanalysis. PLoS Genet 2(12): e190. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1371/journal.pgen.0020190")}.

Examples

set.seed(1)

a <- matrix(0, 93, 170)
p <- 0.2
a[] <- rbinom(length(a), 2, p)
X <- add_code256(big_copy(a, type = "raw"), code = c(0, 1, 2, rep(NA, 253)))
X.svd <- big_SVD(X, fun.scaling = snp_scaleBinom())
str(X.svd)
plot(X.svd$center)
abline(h = 2 * p, col = "red")
plot(X.svd$scale)
abline(h = sqrt(2 * p * (1 - p)), col = "red")

bigsnpr documentation built on Sept. 30, 2024, 9:18 a.m.