popkin popkin

The popkin (“population kinship”) R package estimates the kinship matrix of individuals and FST from their biallelic genotypes. Our estimation framework is the first to be practically unbiased under arbitrary population structures.


The stable version of the package is now on CRAN and can be installed using


The current development version can be installed from the GitHub repository using devtools:

install.packages("devtools") # if needed
install_github('StoreyLab/popkin', build_opts = c())

You can see the package vignette, which has more detailed documentation, by typing this into your R session:



Input data

The examples below assume the following R data variables are present for n individuals and m loci:

The subpops vector is not required, but its use is recommended to improve estimation of the baseline kinship value treated as zero.

If your data is in BED format, popkin will process it efficiently using BEDMatrix. If file is the path to the BED file (excluding .bed extension):

X <- BEDMatrix(file) # load genotype matrix object

popkin functions

This is a quick overview of every popkin function, covering estimation and visualization of kinship and FST from a genotype matrix.

First estimate the kinship matrix from the genotypes X. All downstream analysis require kinship, none use X after this

kinship <- popkin(X, subpops) # calculate kinship from X and optional subpop labels

Plot the kinship matrix, marking the subpopulations. Note inbr_diag replaces the diagonal of kinship with inbreeding coefficients

plot_popkin( inbr_diag(kinship), labs = subpops )

Extract inbreeding coefficients from kinship

inbreeding <- inbr(kinship)

Estimate FST

weights <- weights_subpops(subpops) # weigh individuals so subpopulations are balanced
Fst <- fst(kinship, weights) # use kinship matrix and weights to calculate fst
Fst <- fst(inbreeding, weights) # estimate more directly from inbreeding vector (same result)

Estimate and visualize the pairwise FST matrix

pairwise_fst <- pwfst(kinship) # estimated matrix
leg_title <- expression(paste('Pairwise ', F[ST])) # fancy legend label
plot_popkin(pairwise_fst, labs = subpops, leg_title = leg_title) # NOTE no need for inbr_diag() here!

Rescale the kinship matrix using different subpopulations (implicitly changes the most recent common ancestor population used as reference)

kinship2 <- rescale_popkin(kinship, subpops2)

Please see the popkin R vignette for a description of the key parameters and more detailed examples, including complex plots with multiple kinship matrices and multi-level subpopulation labeling.


Alejandro Ochoa, John D Storey. 2021. “Estimating FST and kinship for arbitrary population structures.” PLoS Genet 17(1): e1009241. PubMed ID 33465078. doi:10.1371/journal.pgen.1009241. bioRxiv doi:10.1101/083923 2016-10-27.

Alejandro Ochoa, John D Storey. 2019. “New kinship and FST estimates reveal higher levels of differentiation in the global human population.” bioRxiv doi:10.1101/653279.

Alejandro Ochoa, John D Storey. 2016. “FST And Kinship for Arbitrary Population Structures I: Generalized Definitions.” bioRxiv doi:10.1101/083915.