Extract per-locus allele frequencies from a filtered table

Computes the frequency of allele 1 at every locus in genome_meta from the allele copies a filtered tidybreed_table selects. The table's identity says what kind of thing is being selected; the user's dplyr::filter() says which ones:

Piped table	Meaning	Copies counted
founder_haplotypes	the founder pool	the filtered pool rows
ind_haplotype	these allele copies	the filtered haplotype rows
any other table with id_ind	these individuals	every ind_haplotype row of the distinct selected id_ind

The frequency is computed in one SQL statement with the filter rendered as a subquery; nothing but the per-locus result is collected into R. For an id_ind table the frequency depends only on which individuals are selected, never on how many rows each has — ind_phenotype with five records per animal gives the same answer as ind_meta for the same animals.

This is the single place tidybreed turns a population selection into p. define_additive_effects() and define_genome_effect_terms() both call it for their base_tbl, so a base selection means the same population in either. Users call it to obtain p for ad_terms().

Usage

extract_allele_freq(tbl)

Arguments

tbl

A tidybreed_table from get_table(), optionally filtered. One of the three shapes above. Columns removed with dplyr::select() are checked before any SQL runs.

Value

A tibble with exactly one row per genome_meta locus in ascending locus_id order: locus_id (integer), locus_name (character), allele_freq (double). allele_freq is NA at a locus the selection has no copies for; it is never 0 in that case, and an observed frequency of exactly 0 or 1 is a real value and is kept. Errors if no locus has a copy at all. Never warns and never writes.

Two notions of line

ind_meta.line_name is a pedigree label — an F1 is whatever it was called; ind_haplotype.line_origin is the founding line each allele copy traces to, carried through every add_offspring() call. They coincide for purebreds and diverge for crosses. get_table(pop, "ind_meta") |> filter(line_name == "F1") counts every copy those animals carry, whichever line it came from; get_table(pop, "ind_haplotype") |> filter(line_origin == "Duroc") counts Duroc copies wherever they sit, at any cross depth.

See also

define_additive_effects(), define_genome_effect_terms(), ad_terms().

Examples

if (FALSE) { # \dontrun{
# Founder pool of one line
p <- pop |> get_table("founder_haplotypes") |>
  dplyr::filter(line_name == "Duroc") |> extract_allele_freq()

# Generation-0 animals
p <- pop |> get_table("ind_meta") |> dplyr::filter(gen == 0L) |>
  extract_allele_freq()

# Duroc copies wherever they sit, including inside crossbreds
p <- pop |> get_table("ind_haplotype") |>
  dplyr::filter(line_origin == "Duroc") |> extract_allele_freq()

# Feed p to ad_terms()
loci <- c("Locus_10", "Locus_44")
tt <- ad_terms(loci, a = c(0.4, 0.1), d = c(0.2, 0.05),
               p = p$allele_freq[match(loci, p$locus_name)],
               coding = "cockerham")
} # }