tests/testthat/test_condense_intsites.R

In cnobles/gintools: Genomic DNA Integration Analysis Tools

context("Condense intsites")

# Generated by gintools:::generate_test_granges()
input_gr <- GenomicRanges::GRanges(
  seqnames = S4Vectors::Rle(factor("chr1"), 100),
  ranges = IRanges::IRanges(
    start = c(
      598L, 598L, 599L, 599L, 599L, 599L, 599L, 599L, 600L, 600L, 600L, 600L, 
      600L, 600L, 600L, 600L, 601L, 601L, 601L, 602L, 657L, 657L, 658L, 658L, 
      658L, 659L, 659L, 659L, 659L, 659L, 659L, 659L, 660L, 660L, 660L, 660L, 
      661L, 661L, 661L, 661L, 743L, 743L, 743L, 744L, 744L, 744L, 744L, 745L, 
      745L, 745L, 745L, 745L, 746L, 746L, 746L, 746L, 746L, 746L, 747L, 747L, 
      780L, 781L, 781L, 781L, 781L, 781L, 781L, 782L, 782L, 782L, 782L, 782L, 
      783L, 783L, 783L, 783L, 783L, 783L, 784L, 784L, 834L, 835L, 835L, 835L, 
      835L, 835L, 835L, 836L, 836L, 837L, 837L, 837L, 837L, 837L, 837L, 837L, 
      837L, 837L, 838L, 838L
    ),
    width = c(
      89L, 99L, 32L, 32L, 41L, 58L, 60L, 95L, 33L, 34L, 39L, 47L, 65L, 72L, 76L,
      77L, 55L, 64L, 85L, 45L, 55L, 99L, 32L, 33L, 58L, 45L, 47L, 72L, 76L, 77L, 
      85L, 95L, 39L, 60L, 65L, 89L, 32L, 34L, 41L, 64L, 41L, 58L, 99L, 32L, 55L, 
      72L, 95L, 32L, 34L, 39L, 64L, 76L, 45L, 60L, 65L, 77L, 85L, 89L, 33L, 47L, 
      89L, 32L, 34L, 41L, 55L, 60L, 72L, 32L, 33L, 64L, 85L, 99L, 45L, 47L, 58L, 
      65L, 76L, 77L, 39L, 95L, 45L, 47L, 55L, 60L, 64L, 76L, 89L, 58L, 72L, 32L,
      34L, 39L, 41L, 65L, 77L, 85L, 95L, 99L, 32L, 33L
    )
  ),
  strand = S4Vectors::Rle(factor("+", levels = c("+", "-", "*")), lengths = 100)
)

output_gr <- GenomicRanges::GRanges(
  seqnames = factor("chr1"),
  ranges = IRanges::IRanges(
    start = c(
      598L, 599L, 600L, 601L, 602L, 657L, 658L, 659L, 660L, 661L, 
      743L, 744L, 745L, 746L, 747L, 780L, 781L, 782L, 783L, 784L, 
      834L, 835L, 836L, 837L, 838L
    ), 
    width = 1L
  ),
  strand = factor("+", levels = c("+", "-", "*")),
  estAbund = c(
    2, 5, 8, 3, 1, 2, 3, 7, 4, 4, 3, 4, 5, 6, 2, 1, 6, 5, 6, 2, 
    1, 6, 2, 9, 2
  ), 
  relAbund = c(
    0.02, 0.05, 0.08, 0.03, 0.01, 0.02, 0.03, 0.07, 0.04, 0.04,
    0.03, 0.04, 0.05, 0.06, 0.02, 0.01, 0.06, 0.05, 0.06, 0.02, 
    0.01, 0.06, 0.02, 0.09, 0.02
  ), 
  relRank = c(
    22L, 10L, 2L, 16L, 25L, 22L, 16L, 3L, 13L, 13L, 
    16L, 13L, 10L, 7L, 22L, 25L, 7L, 10L, 7L, 22L,
    25L, 7L, 22L, 1L, 22L
  ), 
  posid = c(
    "chr1+598", "chr1+599", "chr1+600", "chr1+601", "chr1+602", 
    "chr1+657", "chr1+658", "chr1+659", "chr1+660", "chr1+661", 
    "chr1+743", "chr1+744", "chr1+745", "chr1+746", "chr1+747", 
    "chr1+780", "chr1+781", "chr1+782", "chr1+783", "chr1+784", 
    "chr1+834", "chr1+835", "chr1+836", "chr1+837", "chr1+838"
  )
)


test_that("Consensed integration sites GRanges output", {
  
  set.seed(1)
  res <- condense_intsites(input_gr, return.abundance = TRUE)
  res <- as.data.frame(res, row.names = NULL)
  res$relAbund <- round(res$relAbund, digits = 2)
  expect_equal(res, as.data.frame(output_gr, row.names = NULL))
  
})