fastDiversity

fastDiversity is an R package for fast, flexible calculation of population genetic diversity statistics from biallelic SNP data.

Many existing tools require specific file formats or data structures, and some are no longer actively maintained. fastDiversity is designed to work directly with any SNP dataset encoded in standard biallelic dosage format (0 (homozygous reference, aa), 1 (heterozygous, aA), 2 (homozygous alternate, AA), with NA for missing data), the direct output of most SNP-calling pipelines.

A key design goal is support for datasets spanning multiple species or genetic groups, where locus filtering and statistics need to be calculated independently per group. This is a common scenario in conservation genomics and comparative population studies that is not well served by existing tools.

Key functions

Function	Description
make_allele_list()	Returns a named list of allele sets per group, for allele sharing analysis
calculate_private_alleles()	Counts private (group-exclusive) and total alleles per group
faststats()	Calculates Ho, He, uHe, Fis, uFis, Fst, and Ar per site, with per-group locus filtering, optional bootstrapped confidence intervals, and optional Hardy-Weinberg equilibrium testing
pi_theta_d_stats()	Calculates pi, theta, S, and D with per-group locus filtering

Installation

# Install remotes if needed
if (!requireNamespace("remotes", quietly = TRUE)) install.packages("remotes")

remotes::install_github("eilishmcmaster/fastDiversity", build_vignettes = TRUE)

library(fastDiversity)

# View tutorial vignette
browseVignettes("fastDiversity")

Quick start

library(fastDiversity)

data(example_gt)    # genotype matrix: samples × loci, values 0/1/2/NA
data(example_meta)  # sample metadata with population and site columns

Allele sharing and private alleles:

allele_list <- make_allele_list(example_gt, example_meta$population)

# Visualise overlap between populations
ggvenn::ggvenn(allele_list)

# Count private and total alleles per population
calculate_private_alleles(allele_list)

Diversity statistics:

basicstats <- faststats(
  gt                     = example_gt,
  genetic_group_variable = example_meta$population,
  site_variable          = example_meta$site,
  minimum_n              = 3,
  minimum_loci           = 50,
  maf                    = 0.05,
  max_missingness        = 0.3
)

Returns one row per genetic group × site combination, with columns for Ho, He, uHe, Fis, uFis, Fst, Ar, and the number of individuals and loci used.

With bootstrapped confidence intervals:

basicstats_ci <- faststats(
  gt                     = example_gt,
  genetic_group_variable = example_meta$population,
  site_variable          = example_meta$site,
  minimum_n              = 3,
  minimum_loci           = 50,
  maf                    = 0.05,
  max_missingness        = 0.3,
  get_CI_loci_resampling = TRUE,   # resample loci
  boots                  = 100,
  CI_alpha               = 0.05
)

Individual-level statistics:

# Individual multilocus heterozygosity (MLH)
ind_MLH_ho <- individual_Ho(
  example_gt,
  genetic_group_variable = example_meta$population,
  maf                    = 0.05,
  max_missingness        = 0.3
)

# Individual inbreeding coefficient (Wright's F = 1 - Ho/He)
# He is calculated at the genetic group level, avoiding Wahlund effect inflation
ind_F <- individual_F(
  example_gt,
  genetic_group_variable = example_meta$population,
  maf                    = 0.05,
  max_missingness        = 0.3
)

plot(ind_MLH_ho, ind_F)

Both functions apply MAF and missingness filtering per genetic group before calculation, so individual estimates are based on the same loci used in population-level analyses.

Neutrality statistics:

# Use all available individuals per population
D_n_all <- pi_theta_d_stats(
  example_gt,
  genetic_group_variable = example_meta$population,
  site_variable          = example_meta$population,
  n                      = NULL, # if set to NULL, all individuals are used, but downsampling can be specified
  minimum_n              = 5, 
  minimum_loci           = 5,
  max_missingness        = 0.1,
  maf                    = 0,
  allele_count_min       = 0
)

Returns a table containing nucleotide diversity (pi), segregating sites (S), Wu & Watterson's theta (W), and Tajima's D (D) per site_variable grouping.

For a full worked example with plots, see vignette("fastDiversity-intro").

References

Goudet, J. (2005). Hierfstat, a package for r to compute and test hierarchical F-statistics. Molecular Ecology Notes, 5(1), 184–186. https://doi.org/10.1111/j.1471-8286.2004.00828.x

Kalinowski, S. T. (2004). hp-rare 1.0: a computer program for performing rarefaction on measures of allelic richness. Molecular Ecology Notes, 5(1), 187–189. https://doi.org/10.1111/j.1471-8286.2004.00845.x

Keenan, K. G., McGinnity, P., Cross, T. F., Crozier, W. W., & Prodöhl, P. A. (2013). diveRsity: An R package for the estimation and exploration of population genetics parameters and their associated errors. Methods in Ecology and Evolution, 4(8), 782–788. https://doi.org/10.1111/2041-210x.12067

Korunes, K. L., & Samuk, K. (2021). pixy: Unbiased estimation of nucleotide diversity and divergence in the presence of missing data. Molecular Ecology Resources, 21(4), 1359–1368. https://doi.org/10.1111/1755-0998.13326

Nei, M. (1978). Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics, 89(3), 583–590. https://doi.org/10.1093/genetics/89.3.583