Make ROC curve from sensitivity and specificity

Source: R/gendist_roc_format.R

This is a wrapper function that takes output from the gendist_sensspec_cutoff() function and constructs values for the Receiver Operating Characteristic (ROC) curve

gendist_roc_format(
  cutoff,
  mut_rate,
  mean_gens_pdf,
  max_link_gens = 1,
  max_gens = NULL,
  max_dist = NULL
)

Arguments

cutoff

the maximum genetic distance at which to consider cases linked

mut_rate

mean number of mutations per generation, assumed to be Poisson distributed

mean_gens_pdf

the density distribution of the mean number of generations between cases; the index of this vector is assumed to be the discrete distance between cases

max_link_gens

the maximum generations of separation for linked pairs

max_gens

the maximum number of generations to consider, if NULL (default) value set to the highest number of generations in mean_gens_pdf with a non-zero probability

max_dist

the maximum distance to calculate, if NULL (default) value set to max_gens * 99.9th percentile of mut_rate Poisson distribution

Value

data frame with cutoff, sensitivity, and 1-specificity

See also

Other genetic distance functions: gendist_distribution(), gendist_sensspec_cutoff()

Other ROC functions: optim_roc_threshold()

Author

Shirlee Wohl and Justin Lessler

Examples

# ebola-like pathogen
R <- 1.5
mut_rate <- 1

# use simulated generation distributions
data('genDistSim')
mean_gens_pdf <- as.numeric(genDistSim[genDistSim$R == R, -(1:2)])

# get theoretical genetic distance dist based on mutation rate and generation parameters
dists <- as.data.frame(gendist_distribution(mut_rate = mut_rate,
                       mean_gens_pdf = mean_gens_pdf,
                       max_link_gens = 1))

dists <- reshape2::melt(dists,
                        id.vars = 'dist',
                        variable.name = 'status',
                        value.name = 'prob')

# get sensitivity and specificity using the same paramters
roc_calc <- gendist_roc_format(cutoff = 1:(max(dists$dist)-1),
                               mut_rate = mut_rate,
                               mean_gens_pdf = mean_gens_pdf)