readResults: Reading results of enrichment analysis

In GSEABenchmarkeR: Reproducible GSEA Benchmarking

Description

These functions read results obtained from the application of enrichment methods to multiple datasets for subsequent assessment.

Usage

readResults(
  data.dir,
  data.ids,
  methods,
  type = c("runtime", "ranking", "typeI")
)

Arguments

data.dir	Character. The data directory where results have been saved to.
data.ids	A character vector of dataset IDs.
methods	Methods for enrichment analysis. A character vector with method names typically chosen from sbeaMethods and nbeaMethods, or user-defined functions implementing methods for enrichment analysis.
type	Character. Type of the result. Should be one out of 'runtime', 'ranking', or 'typeI'.

Value

A result list with an entry for each method applied. Each entry stores corresponding runtimes (type="runtime" ), gene set rankings (type="ranking"), or type I error rates (type="typeI") as obtained from applying the respective method to the given datasets.

Author(s)

Ludwig Geistlinger <[e-mail]>

See Also

runEA to apply enrichment methods to multiple datasets.

Examples

    # simulated setup: 
    # 1 methods & 1 datasets
    methods <- paste0("m", 1:2)
    data.ids <- paste0("d", 1:2)

    # result directory
    res.dir <- tempdir()
    sdirs <- file.path(res.dir, methods)
    for(d in sdirs) dir.create(d)
    
    # store runtime & rankings 
    for(m in 1:2)
    {
        rt <- runif(5, min=m, max=m+1)
        for(d in 1:2)
        {
            # runtime
            out.file <- paste(data.ids[d], "txt", sep=".")
            out.file <- file.path(sdirs[m], out.file)
            cat(rt[d], file=out.file) 

            # ranking
            out.file <- sub("txt$", "rds", out.file)
            r <- EnrichmentBrowser::makeExampleData("ea.res") 
            r <- EnrichmentBrowser::gsRanking(r, signif.only=FALSE)
            saveRDS(r, file=out.file)   
        }
    }

    # reading runtime & rankings
    rts <- readResults(res.dir, data.ids, methods, type="runtime")
    rkgs <- readResults(res.dir, data.ids, methods, type="ranking")

GSEABenchmarkeR documentation built on Dec. 12, 2020, 2 a.m.