Calculate false discovery rate of a sample

Source: R/falsediscoveryrate.R

[Deprecated] This function calculates the false discovery rate (proportion of linked pairs that are false positives) in a sample given the sensitivity \(\eta\) and specificity \(\chi\) of the linkage criteria, and sample size \(M\). Assumptions about transmission and linkage (single or multiple) can be specified.

falsediscoveryrate(eta, chi, rho, M, R = NULL, assumption = "mtml")

Arguments

eta

scalar or vector giving the sensitivity of the linkage criteria

chi

scalar or vector giving the specificity of the linkage criteria

rho

scalar or vector giving the proportion of the final outbreak size that is sampled

M

scalar or vector giving the number of cases sampled

R

scalar or vector giving the effective reproductive number of the pathogen (default=NULL)

assumption

a character vector indicating which assumptions about transmission and linkage criteria. Default = 'mtml'. Accepted arguments are:

  1. 'stsl' for the single-transmission single-linkage assumption (prob_trans_stsl()).

  2. 'mtsl' for the multiple-transmission single-linkage assumption (prob_trans_mtsl()).

  3. 'mtml' for the multiple-transmission multiple-linkage assumption (prob_trans_mtml()).

Value

scalar or vector giving the true discovery rate

See also

Other discovery_rate: truediscoveryrate()

Author

John Giles, Shirlee Wohl, and Justin Lessler

Examples

# The simplest case: single-transmission, single-linkage, and perfect sensitivity
falsediscoveryrate(eta=1, chi=0.9, rho=0.5, M=100, assumption='stsl')
#> Warning: `falsediscoveryrate()` was deprecated in phylosamp 1.0.0.
#> ℹ Please use `translink_fdr()` instead.
#> Warning: The `eta` argument of `falsediscoveryrate()` is deprecated as of phylosamp
#> 1.0.0.
#> ℹ Please use the `sensitivity` argument of `translink_fdr()` instead.
#> Warning: The `chi` argument of `falsediscoveryrate()` is deprecated as of phylosamp
#> 1.0.0.
#> ℹ Please use the `specificity` argument of `translink_fdr()` instead.
#> [1] 0.4999926

# Multiple-transmission and imperfect sensitivity
falsediscoveryrate(eta=0.99, chi=0.9, rho=1, M=50, R=1, assumption='mtsl')
#> [1] 0.1373857

# Small outbreak, larger sampling proportion
falsediscoveryrate(eta=0.99, chi=0.95, rho=1, M=50, R=1, assumption='mtml')
#> [1] 0.5427252

# Large outbreak, small sampling proportion
falsediscoveryrate(eta=0.99, chi=0.95, rho=0.5, M=1000, R=1, assumption='mtml')
#> [1] 0.9805463