tests/testthat/test_5_chipenrich.R

In chipenrich: Gene Set Enrichment For ChIP-seq Peak Data

context('Test enrich functions')

test_that('Errors from no peaks', {
	ldef_file = system.file('extdata', 'test_ldef_symbol.txt', package = 'chipenrich')
	peak_file = system.file('extdata', 'test.broadPeak', package = 'chipenrich')
	gs_path = system.file('extdata','vignette_genesets.txt', package='chipenrich')

	expect_error(
		suppressWarnings(chipenrich(peaks = peak_file, genome = 'hg19', genesets = gs_path, locusdef = ldef_file, method = 'binomial', qc_plots = F, out_name = NULL, n_cores = 1)),
		'Intersection between peak midpoints and locus definition is empty')

	expect_error(
		suppressWarnings(chipenrich(peaks = peak_file, genome = 'hg19', genesets = gs_path, locusdef = ldef_file, qc_plots = F, out_name = NULL, n_cores = 1)),
		'Intersection between peak midpoints and locus definition is empty')

	expect_error(
		suppressWarnings(polyenrich(peaks = peak_file, genome = 'hg19', genesets = gs_path, locusdef = ldef_file, qc_plots = F, out_name = NULL, n_cores = 1)),
		'Intersection between peak midpoints and locus definition is empty')

	expect_error(
		suppressWarnings(broadenrich(peaks = peak_file, genome = 'hg19', genesets = gs_path, locusdef = ldef_file, qc_plots = F, out_name = NULL, n_cores=1)),
		'Intersection between peaks and locus definition is empty')
})

data(peaks_E2F4, package = 'chipenrich.data')
peaks_E2F4 = subset(peaks_E2F4, peaks_E2F4$chrom == 'chr1')

data(peaks_H3K4me3_GM12878, package = 'chipenrich.data')
peaks_H3K4me3_GM12878 = subset(peaks_H3K4me3_GM12878, peaks_H3K4me3_GM12878$chrom == 'chr1')

gs_path = system.file('extdata','vignette_genesets.txt', package='chipenrich')

test_that('Test chipenrich method without mappability', {
	results = suppressWarnings(chipenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path,
		locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))

	expect_equal(class(results), 'list')
})

test_that('Test chipenrich method with complete randomization', {
	results = suppressWarnings(chipenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path,
		locusdef = "nearest_tss", qc_plots = F, out_name = NULL, randomization = 'complete', n_cores = 1))

	expect_equal(class(results), 'list')
})

test_that('Test chipenrich method with bylength randomization', {
	results = suppressWarnings(chipenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path,
		locusdef = "nearest_tss", qc_plots = F, out_name = NULL, randomization = 'bylength', n_cores = 1))

	expect_equal(class(results), 'list')
})

test_that('Test chipenrich method with bylocation randomization', {
	results = suppressWarnings(chipenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path,
		locusdef = "nearest_tss", qc_plots = F, out_name = NULL, randomization = 'bylocation', n_cores = 1))

	expect_equal(class(results), 'list')
})

test_that('Test chipenrich method with mappability', {
	results = suppressWarnings(chipenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path,
		locusdef = "nearest_tss", mappability = 24, qc_plots = F,
		out_name = NULL,n_cores=1))

	expect_equal(class(results), 'list')
})

test_that('Test chipenrich_slow method', {
	results = suppressWarnings(chipenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path, method = 'chipenrich_slow',
		locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))

	expect_equal(class(results), 'list')
})

test_that('Test chipapprox method', {
    results = suppressWarnings(chipenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path, method = 'chipapprox',
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))

    expect_equal(class(results), 'list')
})

test_that('Test old chipapprox method', {
	results = suppressWarnings(chipenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path, method = 'chipapprox_old',
		locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))

	expect_equal(class(results), 'list')
})

test_that('Test polyenrich method', {
	results = suppressWarnings(polyenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path, method = 'polyenrich',
		locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))

	expect_equal(class(results), 'list')
})

test_that('Test polyenrich_slow method', {
	results = suppressWarnings(polyenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path, method = 'polyenrich_slow',
		locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))

	expect_equal(class(results), 'list')
})

test_that('Test polyapprox method', {
    results = suppressWarnings(polyenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path, method = 'polyapprox',
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))

    expect_equal(class(results), 'list')
})

peaks_plus_sv = peaks_E2F4
peaks_plus_sv$signalValue = runif(nrow(peaks_E2F4))*100+10

test_that('Error from polyenrich weighted method without a signalValue column', {
    expect_error(
    suppressWarnings(polyenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path,  method = 'polyenrich_weighted', weighting = 'signalValue',
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1)),
    'No signalValue column!')
})

test_that('Error from choosing weights but not polyenrich_weighted', {
    expect_error(
    suppressWarnings(polyenrich(peaks = peaks_plus_sv, genome = 'hg19', genesets = gs_path,  method = 'polyenrich', weighting = 'signalValue',
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1)),
    'You need to use method = polyenrich_weighted if choosing weighting!')
})

test_that('Error from polyenrich weighted picking both signalValue and logsignalValue', {
    expect_error(
    suppressWarnings(polyenrich(peaks = peaks_plus_sv, genome = 'hg19', genesets = gs_path,  method = 'polyenrich_weighted', weighting = c('signalValue','logsignalValue'),
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1)),
    'You can only choose one of signalValue and logsignalValue!')
})

test_that('Error from polyenrich weighted picking no weighting option', {
    expect_error(
    suppressWarnings(polyenrich(peaks = peaks_plus_sv, genome = 'hg19', genesets = gs_path,  method = 'polyenrich_weighted', weighting = NULL,
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1)),
    'No weight options selected!')
})

test_that('Error from polyenrich weighted picking unsupported weights', {
    expect_error(
    suppressWarnings(polyenrich(peaks = peaks_plus_sv, genome = 'hg19', genesets = gs_path,  method = 'polyenrich_weighted', weighting = c('foo','bar'),
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1)),
    'Unsupported weights: foo, bar')
})

test_that('Test polyenrich weighted method with signalValue', {
    results = suppressWarnings(polyenrich(peaks = peaks_plus_sv, genome = 'hg19', genesets = gs_path, method = 'polyenrich_weighted', weighting = 'signalValue',
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))

    expect_equal(class(results), 'list')
})

test_that('Test polyenrich weighted method with logsignalValue', {
    results = suppressWarnings(polyenrich(peaks = peaks_plus_sv, genome = 'hg19', genesets = gs_path, method = 'polyenrich_weighted', weighting = 'logsignalValue',
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))

    expect_equal(class(results), 'list')
})

test_that('Test polyenrich weighted method with multiAssign', {
    results = suppressWarnings(polyenrich(peaks = peaks_plus_sv, genome = 'hg19', genesets = gs_path, method = 'polyenrich_weighted', weighting = 'multiAssign',
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))

    expect_equal(class(results), 'list')
})

test_that('Test polyenrich weighted method with signalValue AND multiAssign', {
    results = suppressWarnings(polyenrich(peaks = peaks_plus_sv, genome = 'hg19', genesets = gs_path, method = 'polyenrich_weighted', weighting = c('signalValue','multiAssign'),
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))

    expect_equal(class(results), 'list')
})

test_that('Test broadenrich method', {
	results = suppressWarnings(broadenrich(peaks = peaks_H3K4me3_GM12878, genome = 'hg19', genesets = gs_path,
		locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores=1))

	expect_equal(class(results), 'list')
})

test_that('Test FET method', {
	results = suppressWarnings(chipenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path, locusdef = "5kb",
		method = "fet", fisher_alt = "two.sided", qc_plots = F, out_name = NULL))

	expect_equal(class(results), 'list')
})

test_that('Test binomial method', {
	results = suppressWarnings(chipenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path, locusdef = "5kb",
		method = "binomial", qc_plots = F, out_name = NULL))

	expect_equal(class(results), 'list')
})

test_that('Test hybrid from scratch', {
    resultsH = suppressWarnings(hybridenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path,
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))

    expect_equal(class(resultsH), 'list')
})

test_that('Test hybrid from inputting two different types of results', {
    resultsC = suppressWarnings(chipenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path,
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))

    resultsP = suppressWarnings(polyenrich(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path,
    locusdef = "nearest_tss", qc_plots = F, out_name = NULL, n_cores = 1))


    resultsH1 = suppressWarnings(hybrid.join(resultsC, resultsP$results))
    resultsH2 = suppressWarnings(hybrid.join(resultsC$results, resultsP))
})

test_that('Test hybrid.join from inputting invalid results', {
    dummyresults = data.frame('P.value' = 1:5/10, 'Geneset.ID' = 5:9, 'foo' = 5:1)

    expect_error(
        suppressWarnings(hybrid.join(1:5, 2:6)),
        'First object is not a valid output or does not have P.value column'
    )

    expect_error(
        suppressWarnings(hybrid.join(dummyresults[,-2], dummyresults)),
        'First object does not have Geneset.ID column'
    )

    expect_error(
        suppressWarnings(hybrid.join(dummyresults[1:2,], dummyresults[3:5,])),
        'No common genesets in the two datasets!'
    )


})

test_that('Test proxReg tss method', {
    results = suppressWarnings(proxReg(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path, reglocation = "tss", qc_plots = F, out_name = NULL))
})

test_that('Test proxReg enhancer method', {
    results = suppressWarnings(proxReg(peaks = peaks_E2F4, genome = 'hg19', genesets = gs_path, reglocation = "enhancer", qc_plots = F, out_name = NULL))
})