HLAsim: Simulate allelic dropouts in HLA data

Description Usage Arguments Value See Also Examples

Map observed/expected genotype frequencies to sampled genotypes

1	map_genotype_frequencies(x)

`x`	Sample returned by `inject_errors`.

A gtf_tbl object that can be used as a training/test data set with the slots:

"Class": Has an error been injected in the EAGs of a sample (“event”, E), or not, (“nonevent”, N).
"genotype": The genotype in format: “03:FYKD/04:ADCGE”.
"zygosity": One of ‘heterozygous’ or ‘homozygous’.
"conc": Randomly attributed DNA concentrations.
"pexp": The expected genotype frequency.
"pobs": The observed genotype frequency.
"log_pexp": The log of the expected genotype frequency.
"log_pobs": The log of the observed genotype frequency.
"log_fold_diff": The log-fold difference between observed and expected genotype frequency.

and the attributes:

"penv": The enclosing environment of the <geno_table> constructor containing memoised mapping and remapping functions created by make_mapper, as well as genotype frequencies calculated by gtf.
"breaks": The cut points used to slice up the DNA concentration.

Other simulation.functions: inject_errors, make_genotype_sampler

## Not run: 
## Extract HLA-DPB1 genotype frequencies
dpb1 <- HLA("DPB1", "01/01/2014", "23/03/2015")

## Restrict the data to the German sample
dpb1.de <- dpb1[provenance == "DE"]

## Generate an EAG table
dpb1_eag1412 <- eag_table(gene = "DPB1", nextype_basis_id = "1412")

## Generate a distribution of DNA concentrations
conc <- sample_dna_concentration(dpb1.de, n = 1000, ncores = 8)

## Generate a sampling function
sample_dpb1_de <- make_genotype_sampler(dpb1.de, dpb1_eag1412)

## Load precomputed DNA-concentration-dependent error distribution for HLA-A
data(cdfA)

## Sample genotypes
n <- 10000
bin_size <- 3
perr <- 0.01
odds <- 0.25
ans <- sample_dpb1_de(conc, n, bin_size) %>%
  inject_errors(cdfA, perr, odds) %>%
  map_genotype_frequencies()
ans

## End(Not run)