Define the genome structure of a breeding population

Adds genome tables (genome_meta, genome_map, ind_haplotype, ind_genotype, chr_inheritance, chr_recombination) to a population opened with open_pop(). Pipe-friendly — accepts a tidybreed_pop and returns a tidybreed_pop. Physical position (pos_bp, base pairs) lives in genome_meta; the genetic map (pos_cM, centiMorgans) lives in the long genome_map table. Haplotypes are stored in long format; ind_genotype is an on-demand dosage cache populated by add_dosage().

This is the second step in setting up a simulation:

pop <- open_pop(...) |>
  define_genome(n_loci = 50000, n_chr = 18, chr_len_Mb = 100)

To create the founder haplotype pool needed by add_founders(), call define_founder_haplotypes() after this function.

The call is atomic: every table is created inside a single transaction, so an error at any point (invalid argument, impossible locus spacing, failed validation) leaves the database exactly as it was and the population can be re-used for a corrected call.

Usage

define_genome(
  pop,
  n_loci,
  n_chr,
  chr_len_Mb,
  cM_per_Mb = 1,
  locus_names = NULL,
  chr_names = NULL,
  recombines_M = TRUE,
  recombines_F = TRUE
)

Arguments

pop: A tidybreed_pop object from open_pop(). Must not already have a genome defined — define_genome() is called once per population.
n_loci: Integer scalar. Total number of loci to simulate. Must be a whole number and at least n_chr (every chromosome needs a locus).
n_chr: Integer scalar. Number of chromosomes. Must be a whole number.
chr_len_Mb: Numeric scalar or numeric vector of length n_chr. Chromosome length(s) in megabases. A single scalar applies the same length to all chromosomes; a vector specifies each chromosome separately. Determines the physical coordinate genome_meta.pos_bp.
cM_per_Mb: Numeric scalar or numeric vector of length n_chr. Genetic map rate in centiMorgans per megabase, used to derive the genetic map position genome_map.pos_cM from pos_bp (pos_cM = pos_bp/1e6 * cM_per_Mb). Default 1.0. Must be finite and strictly positive.
locus_names: Character vector of length n_loci or NULL. Custom locus names. When NULL (default), names are auto-generated as "Locus_1", "Locus_2", etc.
chr_names: Character vector of length n_chr or NULL. Custom chromosome names. When NULL (default), chromosomes are numbered 1, 2, ..., n_chr.
recombines_M, recombines_F: Logical scalar. Genome-wide recombination default for male-parent and female-parent meiosis, seeded into chr_recombination. Both default TRUE. Set one FALSE for a genome-wide, whole-genome achiasmatic sex (e.g. recombines_F = FALSE for silkworm females, recombines_M = FALSE for Drosophila males) instead of writing a per-chromosome rule for every chromosome. Per-chromosome overrides (Y, W, Drosophila chr4) still go through define_chromosome().

Value

The input tidybreed_pop with genome tables added.

Examples

if (FALSE) { # \dontrun{
# Simple setup
pop <- open_pop(pop_name = "A", db_name = ":memory:") |>
  define_genome(n_loci = 1000, n_chr = 10, chr_len_Mb = 100)

# Different chromosome lengths (cattle)
pop <- open_pop(pop_name = "Cattle", db_name = ":memory:") |>
  define_genome(
    n_loci = 50000,
    n_chr  = 30,
    chr_len_Mb = c(158, 137, 121, 120, 121, 119, 112, 113, 105, 104,
                   107,  91,  84,  84,  85,  81,  75,  66,  64,  72,
                    71,  61,  52,  62,  42,  51,  45,  46,  51,  42)
  )

# With founder haplotypes
pop <- open_pop(pop_name = "B", db_name = ":memory:") |>
  define_genome(n_loci = 1000, n_chr = 10, chr_len_Mb = 100) |>
  define_founder_haplotypes(n_haplotypes = 100,
                            min_allele_freq = 0.05,
                            max_allele_freq = 0.95)

# Genome-wide achiasmy — Drosophila males do not recombine (all chromosomes).
# Per-chromosome overrides (e.g. chr4) still go through define_chromosome().
pop <- open_pop(pop_name = "fly", db_name = ":memory:") |>
  define_genome(n_loci = 1000, n_chr = 4, chr_len_Mb = 100,
                recombines_M = FALSE)
} # }