README.md

In OchoaLab/popkinsuppl: Supplement to popkin package

popkinsuppl

popkinsuppl is a supplementary package containing custom implementations of the existing kinship and Fst approaches benchmarked in Ochoa and Storey (2016). The main package, popkin, provides our novel estimators and is available on CRAN.

Use of the functions in popkinsuppl is discouraged except for comparisons, which is why we are not planning on making it available on CRAN. This package does have documentation and tests of the key functions.

Installation

You can install popkinsuppl from the GitHub repository using devtools:

install.packages("devtools") # if needed
library(devtools)
install_github('OchoaLab/popkinsuppl')

Example

Simulate random genotypes

First we need a genotype matrix X and, for some of the functions, subpopulation labels labs. Here we construct simple data without population structure. The key functions are only interesting under a structured population, but will work with any data in the correct format.

# dimensions of simulated data
n_ind <- 100
m_loci <- 1000
n_data <- n_ind * m_loci

# missingness rate
miss <- 0.1

# simulate ancestral allele frequencies
# uniform (0,1)
p_anc <- runif(m_loci)

# simulate some binomial data
X <- rbinom(n_data, 2, p_anc)

# sprinkle random missingness
X[ sample(X, n_data * miss) ] <- NA

# turn into a matrix
X <- matrix(X, nrow = m_loci, ncol = n_ind)

# create fake subpopulation labels
# (note there are no actual subpopulations in genotype matrix)
# k_subpops groups of equal size
k_subpops <- 10
labs <- ceiling( (1 : n_ind) / k_subpops )

Standard Kinship estimates

There are two forms of what we call the "Standard Kinship" estimator, both are implemented. The "ratio-of-means" version is more robust, having known convergence properties. However, the "mean-of-ratios" version is much more prevalent in practice and in the literature.

library(popkinsuppl)

# estimate kinship matrices

# ... ratio-of-means version
kinship_rom <- kinship_std(X)

# ... mean-of-ratios version
kinship_mor <- kinship_std(X, mean_of_ratios = TRUE)

FST estimates

This package implements the Weir-Cockerham and Generalized "Hudson" FST estimators. It also provides a function to compute a matrix of pairwise Hudson FST estimates, where every entry compares a pair of subpopulations.

# estimate FST using the Weir-Cockerham formula
fst_wc_obj <- fst_wc(X, labs)
# the genome-wide FST estimate
fst_wc_obj$fst
# vector of per-locus FST estimates
fst_wc_obj$fst_loci

# estimate FST using the Weir-Hill formula
fst_wh_obj <- fst_wh(X, labs)
# the genome-wide FST estimate
fst_wh_obj$fst
# vector of per-locus FST estimates
fst_wh_obj$fst_loci

# estimate FST using the "Hudson" formula
fst_hudson_k_obj <- fst_hudson_k(X, labs)
# the genome-wide FST estimate
fst_hudson_k_obj$fst
# vector of per-locus FST estimates
fst_hudson_k_obj$fst_loci

# estimated pairwise FST matrix using the "Hudson" formula
fst_hudson_matrix <- fst_hudson_pairwise(X, labs)

OchoaLab/popkinsuppl documentation built on May 17, 2022, 9:50 a.m.