Skip to contents

This function calculates the global False Discovery Rate (FDR) as a percentage based on the total number of target and decoy identifications in the data.

Usage

global_fdr(data)

Arguments

data

A data frame containing the identification data. It must include a logical column isDecoy indicating whether each row corresponds to a decoy identification.

Value

A data frame with the total counts of target and decoy identifications, and the global FDR expressed as a percentage.

Examples

# Example usage with a sample dataset
sample_data <- data.frame(
  isDecoy = c(TRUE, FALSE, FALSE, TRUE, FALSE)
)
global_fdr(sample_data)
#> # A tibble: 1 × 3
#>   Target Decoy `Global FDR (%)`
#>    <int> <int>            <dbl>
#> 1      3     2             66.7