Compute per-window reference allele frequencies across populations

Description

Iterates over each population defined in pop.file, splits the genotype data by contig, and slides a fixed-size window along each contig to compute the reference allele frequency within that window. Both diploid genotypes ("0/0", "0/1", "1/1") and haploid genotypes ("0", "1") are recognised. Allele frequency is agnostic to ploidy. The resulting per-window frequencies are the direct input expected by [pairwise.fst()].

Usage

allele.freq.WS(
  geno.data,
  pop.file,
  contigs,
  positions,
  window.size,
  verbose = TRUE
)

Arguments

Value

A [data.table::data.table] with one row per population-contig-window combination and the following columns:

Pop

Population label.

Contig

Contig (chromosome) name.

Window_starts

Genomic coordinate (bp) of the first position in the window.

Window_ends

Genomic coordinate (bp) of the last position in the window (Window_starts + window.size - 1).

N_sites

Total number of called genotype entries (diploid + haploid) within the window.

Freq.A

Frequency of the reference allele in the window, computed as (2*N_AA + N_Aa + N_A) / (2*N_dip + N_hap).

See Also

[pairwise.fst()] for computing Fst from the output of this function.

Examples

vcf_path <- system.file("extdata",
                        "example.vcf",
                        package = "HaploDiploidEquilibrium")
result   <- vcf2GT(vcf_path)
gt       <- result$gt_matrix
contigs  <- result$contig_vector
pos      <- result$positions

pop.file <- data.frame(ID  = colnames(gt),
                       Pop = c("PopA","PopA","PopB","PopB","PopB"))

af <- allele.freq.WS(geno.data  = gt,
                     pop.file   = pop.file,
                     contigs    = contigs,
                     positions  = pos,
                     window.size = 10000)

Compute per-window reference allele frequencies by sex

Description

Iterates over each population defined in pop.file, splits the genotype data by contig, and slides a fixed-size window along each contig to compute the reference allele frequency separately for diploid individuals (females), haploid individuals (males), and both sexes combined. Sex is inferred from ploidy: diploid genotypes ("0/0", "0/1", "1/1") are assumed to belong to females and haploid genotypes ("0", "1") to males.

Usage

compute.Female.Male.allele.W(
  geno.data,
  pop.file,
  contigs,
  positions,
  window.size,
  verbose = TRUE
)

Arguments

Value

A [data.table::data.table] with one row per population-contig-window combination and the following columns:

Pop

Population label.

Contig

Contig (chromosome) name.

Window_starts

Genomic coordinate (bp) of the first position in the window.

Window_ends

Genomic coordinate (bp) of the last position in the window (Window_starts + window.size - 1).

N_sites

Total number of called genotype entries (diploid + haploid) within the window.

Females.freq

Reference allele frequency computed from diploid genotypes only: (2*N_AA + N_Aa) / (2*N_dip).

Males.freq

Reference allele frequency computed from haploid genotypes only: N_A / N_hap.

Total.freq

Reference allele frequency computed from both sexes combined: (2*N_AA + N_Aa + N_A) / (2*N_dip + N_hap).

See Also

[summarize_sex_ref()] for computing weighted genome-wide summary statistics from the output.

Examples

vcf_path <- system.file("extdata",
                        "example.vcf",
                        package = "HaploDiploidEquilibrium")
result <- vcf2GT(vcf_path)
gt       <- result$gt_matrix
contigs  <- result$contig_vector
pos      <- result$positions

pop.file <- data.frame(ID  = colnames(gt),
                       Pop = c("PopA","PopA","PopB","PopB","PopB"))

sex_ref <- compute.Female.Male.allele.W(geno.data   = gt,
                                        pop.file    = pop.file,
                                        contigs     = contigs,
                                        positions   = pos,
                                        window.size = 10000)

Compute per-window Nei's H (gene diversity)

Description

Iterates over each population defined in pop.file, splits the genotype data by contig, and slides a fixed-size window along each contig to compute Nei's H (probability of sampling two different alleles) within that window. Nei's H is calculated as 2pq, where p and q are the reference and alternative allele frequencies respectively. Both diploid genotypes ("0/0", "0/1", "1/1") and haploid genotypes ("0", "1") are recognised when computing allele frequencies. Despite the different ploidies, allele frequencies should be the same between sexes, which means that Nei's H is agnostic to ploidy.

Usage

compute_Hs_W(
  geno.data,
  pop.file,
  contigs,
  positions,
  window.size,
  verbose = TRUE
)

Arguments

Value

A [data.table::data.table] with one row per population-contig-window combination and the following columns:

Pop

Population label.

Contig

Contig (chromosome) name.

Window_starts

Genomic coordinate (bp) of the first position in the window.

Window_ends

Genomic coordinate (bp) of the last position in the window (Window_starts + window.size - 1).

N_sites

Total number of called genotype entries (diploid + haploid) within the window.

Neis_H

Nei's H (gene diversity) for the window, computed as 2 * Freq.Ref * Freq.Alt.

See Also

[summarize_NeisH()] for computing weighted genome-wide summary statistics from the output.

Examples

vcf_path <- system.file("extdata",
                        "example.vcf",
                        package = "HaploDiploidEquilibrium")

result <- vcf2GT(vcf_path)
gt       <- result$gt_matrix
contigs  <- result$contig_vector
pos      <- result$positions

pop.file <- data.frame(ID  = colnames(gt),
                       Pop = c("PopA","PopA","PopB","PopB","PopB"))

hs <- compute_Hs_W(geno.data   = gt,
                   pop.file    = pop.file,
                   contigs     = contigs,
                   positions   = pos,
                   window.size = 10000)

Compute per-window genotype frequencies, allele frequencies, and Fis

Description

Iterates over each population defined in pop.file, splits the genotype data by contig, and slides a fixed-size window along each contig to compute observed and expected genotype frequencies, allele frequencies, and the inbreeding coefficient (Fis). Expected genotype frequencies are derived from the haplo-diploid equilibrium model, where the proportion of diploid and haploid individuals in the population is controlled by dip_freq. Sex is inferred from ploidy: diploid genotypes ("0/0", "0/1", "1/1") are assumed to belong to females and haploid genotypes ("0", "1") to males. Fis is computed as 1 - (Obs.Het / Exp.Het) and is set to NA when expected heterozygosity is zero.

Usage

compute_allele.freqs_W(
  geno.data,
  pop.file,
  contigs,
  positions,
  window.size,
  dip_freq,
  verbose = TRUE
)

Arguments

Details

Haplo-diploid equilibrium model assumes a equal sex-ratio for three main reasons: 1: A sex-ratio different from 1:1 (e.g. 1:2) would break the assumptions of equal probability of contributing to next generation (selection and even drift) 2: Population sex-ratio is not sample sex-ratio. Sample sex-ratio might be different from population's sex-ratio due to time of the year when sampling took place, flower resources, etc. 3: Sex-ratio might be different from didderent populations, across time and evolving. So its estimation for one population might not hold for another or the same in the next breeding season

For the above mentioned reasons, we highly recomend that the sex-ratio is left has 0.5 (1:1) unless strong evidence existis of otherwise. A true sex-ratio different from 1:1 (assumed) should also be considered to explain the possible differences between observed and expected genotype frequencies. In other words, sex-ratio different from 1:1 should change the genotype frequencies but not the allele frequencies.

Fis might not be reliable when the number of sampled diploids is to low

Value

A [data.table::data.table] with one row per population-contig-window combination and the following columns:

Pop

Population label.

Contig

Contig (chromosome) name.

Window_starts

Genomic coordinate (bp) of the first position in the window.

Window_ends

Genomic coordinate (bp) of the last position in the window (Window_starts + window.size - 1).

N_sites

Total number of called genotype entries (diploid + haploid) within the window.

N_F

Number of diploid (female) genotype entries in the window.

N_M

Number of haploid (male) genotype entries in the window.

N_AA

Count of homozygous reference (AA) genotypes.

N_Aa

Count of heterozygous (Aa) genotypes.

N_aa

Count of homozygous alternative (aa) genotypes.

N_A

Count of haploid reference (A) genotypes.

N_a

Count of haploid alternative (a) genotypes.

Freq.Ref

Overall reference allele frequency in the window.

Freq.Alt

Overall alternative allele frequency in the window.

Obs.Hom

Observed proportion of homozygous diploid genotypes (AA + aa) relative to total entries.

Obs.Het

Observed proportion of heterozygous diploid genotypes (Aa) relative to total entries.

Obs.M.Ref

Observed proportion of haploid reference genotypes (A) relative to total entries.

Exp.Hom

Expected frequency of homozygous diploid genotypes under the haplo-diploid equilibrium model.

Exp.Het

Expected frequency of heterozygous diploid genotypes under the haplo-diploid equilibrium model.

Exp.M.Ref

Expected frequency of haploid reference genotypes under the haplo-diploid equilibrium model.

Fis

Inbreeding coefficient for the window, or NA when expected heterozygosity is zero.

See Also

[summarize_geno()] for computing weighted genome-wide summary statistics from the output.

Examples

vcf_path <- system.file("extdata",
                        "example.vcf",
                        package = "HaploDiploidEquilibrium")

result <- vcf2GT(vcf_path)
gt       <- result$gt_matrix
contigs  <- result$contig_vector
pos      <- result$positions

pop.file <- data.frame(ID  = colnames(gt),
                       Pop = c("PopA","PopA","PopB","PopB","PopB"))

geno <- compute_allele.freqs_W(geno.data   = gt,
                               pop.file    = pop.file,
                               contigs     = contigs,
                               positions   = pos,
                               window.size = 10000,
                               dip_freq    = 0.5)

Compute pairwise Fst for all population pairs

Description

Takes the per-window allele frequency table produced by [allele.freq.WS()] and computes Wright's Fst for every unique pair of populations. Within each contig-window, Fst is estimated as the ratio of the among-population variance in reference allele frequency to its expected maximum under panmixia: Fst = Var(p) / (mean(p) * (1 - mean(p))). Windows in which the mean reference allele frequency is 0 or 1 (i.e. monomorphic across the pair) are set to 0.

Usage

pairwise.fst(allele.freq.table, verbose = TRUE)

Arguments

Value

A [data.table::data.table] with one row per population-pair-contig- window combination and the following columns:

Contig

Contig (chromosome) name.

window_lims

A character string of the form "Window_starts - Window_ends" identifying the window.

Sum.Sites

Total number of called genotype entries summed across both populations in the window.

Mean.p

Mean reference allele frequency across the two populations in the window.

Var.p

Population variance of the reference allele frequency across the two populations in the window.

Fst

Estimated Fst for the window.

Pop_pair

A character string of the form "Pop1 - Pop2" identifying the population pair.

See Also

[allele.freq.WS()] for computing the input allele frequency table, and [summarize_fst()] for computing weighted genome-wide summary statistics from the output.

Examples

vcf_path <- system.file("extdata",
                        "example.vcf",
                        package = "HaploDiploidEquilibrium")
result   <- vcf2GT(vcf_path)
gt       <- result$gt_matrix
contigs  <- result$contig_vector
pos      <- result$positions

pop.file <- data.frame(ID  = colnames(gt),
                       Pop = c("PopA","PopA","PopB","PopB","PopB"))

af <- allele.freq.WS(geno.data  = gt,
                     pop.file   = pop.file,
                     contigs    = contigs,
                     positions  = pos,
                     window.size = 10000)
fst <- pairwise.fst(af)

Summarize per-window Nei's H per population

Description

Computes the site-count-weighted mean and standard deviation of Nei's H across all windows for each population, using the per-window table produced by [compute_Hs_W()]. Weighting by N_sites ensures that windows with more called genotypes contribute more to the estimate.

Usage

summarize_NeisH(neis_table)

Arguments

Value

A [tibble::tibble] with one row per population and the following columns:

Pop

Population label.

wMean.Neis_H

Weighted mean of Nei's H across all windows.

wSD.Neis_H

Weighted standard deviation of Nei's H across all windows.

See Also

[compute_Hs_W()] for computing the input per-window table.

Examples

vcf_path <- system.file("extdata",
                        "example.vcf",
                        package = "HaploDiploidEquilibrium")

result <- vcf2GT(vcf_path)
gt       <- result$gt_matrix
contigs  <- result$contig_vector
pos      <- result$positions

pop.file <- data.frame(ID  = colnames(gt),
                       Pop = c("PopA","PopA","PopB","PopB","PopB"))

hs <- compute_Hs_W(geno.data   = gt,
                   pop.file    = pop.file,
                   contigs     = contigs,
                   positions   = pos,
                   window.size = 10000)
summary <- summarize_NeisH(hs)

Summarize genome-wide Fst per population pair

Description

Computes the site-count-weighted mean and standard deviation of Fst across all windows for each population pair, using the per-window Fst table produced by [pairwise.fst()]. Weighting by Sum.Sites ensures that windows with more called genotypes contribute more to the estimate.

Usage

summarize_fst(fst_table)

Arguments

Value

A [tibble::tibble] with one row per population pair and the following columns:

Pop_pair

A character string identifying the population pair.

wMean.Fst

Weighted mean of Fst across all windows.

wSD.Fst

Weighted standard deviation of Fst across all windows.

See Also

[pairwise.fst()] for computing the input per-window Fst table.

Examples

vcf_path <- system.file("extdata",
                        "example.vcf",
                        package = "HaploDiploidEquilibrium")
result   <- vcf2GT(vcf_path)
gt       <- result$gt_matrix
contigs  <- result$contig_vector
pos      <- result$positions

pop.file <- data.frame(ID  = colnames(gt),
                       Pop = c("PopA","PopA","PopB","PopB","PopB"))

af <- allele.freq.WS(geno.data  = gt,
                     pop.file   = pop.file,
                     contigs    = contigs,
                     positions  = pos,
                     window.size = 10000)
fst <- pairwise.fst(af)
summary <- summarize_fst(fst)

Summarize per-window genotype frequencies per population

Description

Computes the site-count-weighted mean and standard deviation of observed and expected heterozygosity, and of observed and expected haploid reference allele frequency, across all windows for each population. Uses the per-window table produced by [compute_allele.freqs_W()]. Weighting by N_sites ensures that windows with more called genotypes contribute more to each estimate.

Usage

summarize_geno(geno_table)

Arguments

Value

A [tibble::tibble] with one row per population and the following columns:

Pop

Population label.

wMean.Exp.Het

Weighted mean of expected heterozygosity across all windows.

wSD.Exp.Het

Weighted standard deviation of expected heterozygosity across all windows.

wMean.Obs.Het

Weighted mean of observed heterozygosity across all windows.

wSD.Obs.Het

Weighted standard deviation of observed heterozygosity across all windows.

wMean.Exp.M.Ref

Weighted mean of expected haploid reference allele frequency across all windows.

wSD.Exp.M.Ref

Weighted standard deviation of expected haploid reference allele frequency across all windows.

wMean.Obs.M.Ref

Weighted mean of observed haploid reference allele frequency across all windows.

wSD.Obs.M.Ref

Weighted standard deviation of observed haploid reference allele frequency across all windows.

See Also

[compute_allele.freqs_W()] for computing the input per-window table.

Examples

vcf_path <- system.file("extdata",
                        "example.vcf",
                        package = "HaploDiploidEquilibrium")

result <- vcf2GT(vcf_path)
gt       <- result$gt_matrix
contigs  <- result$contig_vector
pos      <- result$positions

pop.file <- data.frame(ID  = colnames(gt),
                       Pop = c("PopA","PopA","PopB","PopB","PopB"))

geno <- compute_allele.freqs_W(geno.data   = gt,
                               pop.file    = pop.file,
                               contigs     = contigs,
                               positions   = pos,
                               window.size = 10000,
                               dip_freq    = 0.5)
summary <- summarize_geno(geno)

Summarize per-sex reference allele frequencies per population

Description

Computes the site-count-weighted mean and standard deviation of the reference allele frequency for females, males, and both sexes combined, across all windows for each population. Uses the per-window table produced by [compute.Female.Male.allele.W()]. Weighting by N_sites ensures that windows with more called genotypes contribute more to each estimate.

Usage

summarize_sex_ref(allele_table)

Arguments

Value

A [tibble::tibble] with one row per population and the following columns:

Pop

Population label.

wMean.F.Ref

Weighted mean of the female reference allele frequency across all windows.

wSD.F.Ref

Weighted standard deviation of the female reference allele frequency across all windows.

wMean.M.Ref

Weighted mean of the male reference allele frequency across all windows.

wSD.M.Ref

Weighted standard deviation of the male reference allele frequency across all windows.

wMean.T.Ref

Weighted mean of the combined reference allele frequency across all windows.

wSD.T.Ref

Weighted standard deviation of the combined reference allele frequency across all windows.

See Also

[compute.Female.Male.allele.W()] for computing the input per-window table.

Examples

vcf_path <- system.file("extdata",
                        "example.vcf",
                        package = "HaploDiploidEquilibrium")
result <- vcf2GT(vcf_path)
gt       <- result$gt_matrix
contigs  <- result$contig_vector
pos      <- result$positions

pop.file <- data.frame(ID  = colnames(gt),
                       Pop = c("PopA","PopA","PopB","PopB","PopB"))

sex_ref <- compute.Female.Male.allele.W(geno.data   = gt,
                                        pop.file    = pop.file,
                                        contigs     = contigs,
                                        positions   = pos,
                                        window.size = 10000)
summary <- summarize_sex_ref(sex_ref)

Import a VCF file and extract genotype and positional data

Description

Reads a VCF file from disk and extracts three pieces of information: the genotype calls (the GT field), the contig (chromosome) name for each variant site, and its physical position. These are the three inputs required by the windowed summary-statistic functions in this package.

Usage

vcf2GT(path_to_vcf)

Arguments

Value

A named list with three elements:

contig_vector

A character vector of length n_sites containing the contig (chromosome) name for each variant.

positions

A numeric vector of length n_sites containing the physical position (bp) of each variant.

gt_matrix

A character matrix of genotype strings with dimensions n_sites x n_individuals (e.g. "0/0", "0/1", "1"). Row names are inherited from the VCF variant records and column names correspond to sample identifiers in the VCF header.

See Also

[vcfR::read.vcfR()] for full control over VCF parsing options.

Examples

vcf_path <- system.file("extdata",
                        "example.vcf",
                        package = "HaploDiploidEquilibrium")

result <- vcf2GT(vcf_path)

head(result$contig_vector)
head(result$positions)
head(result$gt_matrix)