Get the oracle misclassification error rates (conditional on true genotype) directly from the joint distribution of the genotype and the oracle estimator.

Get the oracle misclassification error rates (conditional on true genotype) directly from the joint distribution of the genotype and the oracle estimator.

Usage

oracle_mis_vec_from_joint(jd)

Arguments

jd

A matrix of numerics. Element (i, j) is the probability of genotype i - 1 and estimated genotype j - 1. This is usually obtained from oracle_joint.

Value

A vector of numerics. Element i is the oracle misclassification error rate when genotyping an individual with actual genotype i + 1.

References

• Gerard, D., Ferrão, L. F. V., Garcia, A. A. F., & Stephens, M. (2018). Genotyping Polyploids from Messy Sequencing Data. Genetics, 210(3), 789-807. doi:10.1534/genetics.118.301468 .

See also

oracle_joint for getting jd. oracle_mis_vec for not having to first calculate jd.

Author

David Gerard

Examples

## Hardy-Weinberg population with allele-frequency of 0.75.
## Moderate bias and moderate overdispersion.
ploidy <- 6
dist <- stats::dbinom(0:ploidy, ploidy, 0.75)
jd <- oracle_joint(n = 100, ploidy = ploidy, seq = 0.001,
                   bias = 0.7, od = 0.01, dist = dist)
oracle_mis_vec_from_joint(jd = jd)
#> [1] 0.001855178 0.186231038 0.262779904 0.249400633 0.177957888 0.103565813
#> [7] 0.005097110

## Compare to oracle_cor
oracle_mis_vec(n = 100, ploidy = ploidy, seq = 0.001,
               bias = 0.7, od = 0.01, dist = dist)
#> [1] 0.001855178 0.186231038 0.262779904 0.249400633 0.177957888 0.103565813
#> [7] 0.005097110