Evaluate assumptions of the Target Decoys Approach for multiple search engines

Source: R/Scores.R

Create diagnostic PP plots in one figure to evaluate the TDA assumptions for multiple search engines. The function provides the possibility to evaluate each of the sub-engines and the overall itself.

createPPlotScores(object, scores, decoy, log10 = TRUE)

Arguments

object

A data.frame, mzID or mzRident object.

scores

A character vector of multiple score names from the input object. Typically from different search engines.

decoy

character, name of the variable that indicates if the peptide matches to a target or to a decoy. When this value is missing, a Shiny gadget is launched to select it interactively.

log10

logical to indicate if the score should be -log10-transformed. Default: TRUE. When this value is missing, a Shiny gadget is launched to select it interactively.

Value

One PP plot with all original pi0, and a standardized / rescaled PP plot with all pi0 set to 0.

Author

Elke Debrie, Lieven Clement

Examples

library(mzID)

## Use one of the example files in the mzID package
exampleFile <- system.file("extdata", "55merge_tandem.mzid", package = "mzID")
mzIDexample <- mzID(exampleFile)
#> reading 55merge_tandem.mzid... DONE!

plots <- createPPlotScores(mzIDexample,
    scores = c("x\\!tandem:hyperscore", "x\\!tandem:expect"),
    decoy = "isdecoy", log10 = TRUE
)