Nothing
test_that("geometric_means() sets zeros to 1 and ignores missing probes", {
m <- matrix(c(0, 4, NA, 1, 4, 16), nrow = 3)
expect_equal(NACHO:::geometric_means(m), c(2, 4))
})
negative_matrix <- function(means) {
matrix(
rep(means, times = 2),
ncol = 2,
dimnames = list(sprintf("NEG_%s", LETTERS[seq_along(means)]), c("S1", "S2"))
)
}
exclude_nsolver <- function(means) {
NACHO:::excluded_negatives(
negative_matrix(means),
rep("Negative", length(means)),
"nsolver"
)
}
test_that("the Bruker rule drops at most two negatives 3-fold above the others", {
expect_identical(exclude_nsolver(c(10, 11, 9, 12, 8, 10)), character(0))
expect_identical(exclude_nsolver(c(40, 11, 9, 12, 8, 10)), "NEG_A")
expect_identical(
exclude_nsolver(c(40, 50, 9, 12, 8, 10)),
c("NEG_B", "NEG_A")
)
expect_identical(exclude_nsolver(c(40, 50, 60, 12, 8, 10)), character(0))
expect_identical(exclude_nsolver(c(40, 11, 9)), character(0))
})
test_that("a negative probe with only missing counts is left out of the rule", {
counts <- negative_matrix(c(40, 11, 9, 12))
counts[c("NEG_C", "NEG_D"), ] <- NA
expect_identical(
NACHO:::excluded_negatives(counts, rep("Negative", 4), "nsolver"),
character(0)
)
padded <- negative_matrix(c(40, 11, 9, 12, 10))
padded["NEG_E", ] <- NA
expect_identical(
NACHO:::excluded_negatives(padded, rep("Negative", 5), "nsolver"),
"NEG_A"
)
})
test_that("the legacy rule keeps the NACHO 2 median rule", {
counts <- matrix(
c(10, 10, 10, 30, 10, 11, 9, 31),
ncol = 2,
dimnames = list(1:4, 1:2)
)
expect_identical(
NACHO:::excluded_negatives(counts, rep("Negative", 4), "legacy"),
"4"
)
})
test_that("the legacy rule never reports a missing negative probe name", {
counts <- matrix(
c(10, 10, 10, NA, NA, NA, 10, 11, 9, 30, 31, 29),
nrow = 4,
byrow = TRUE,
dimnames = list(c("A", "B", "C", "D"), 1:3)
)
expect_identical(
NACHO:::excluded_negatives(counts, rep("Negative", 4), "legacy"),
"D"
)
})
test_that("nsolver PCL leaves POS_F out and adds 1 to every count", {
names <- sprintf("POS_%s(%s)", LETTERS[1:6], c(128, 32, 8, 2, 0.5, 0.125))
positives <- matrix(c(1000, 260, 70, 15, 6, 0), ncol = 1)
expected <- stats::cor(
log2(positives[1:5] + 1),
log2(c(128, 32, 8, 2, 0.5))
)^2
expect_equal(
NACHO:::sample_pcl(positives, names, "nsolver"),
round(expected, 5)
)
legacy <- stats::cor(
log2(as.vector(positives) + 1),
log2(c(128, 32, 8, 2, 0.5, 0.125))
)^2
expect_equal(NACHO:::sample_pcl(positives, names, "legacy"), round(legacy, 5))
})
test_that("normalised counts are neither rounded nor floored", {
counts <- matrix(
c(3L, 7L, 11L, 2L),
ncol = 2,
dimnames = list(c("a", "b"), c("S1", "S2"))
)
scaled <- NACHO:::scale_counts(
counts,
background = c(4, 1),
background_mode = "subtract",
positive_factor = c(1.5, 0.5)
)
expect_identical(
scaled,
matrix(c(0, 4.5, 5, 0.5), ncol = 2, dimnames = dimnames(counts))
)
expect_identical(
NACHO:::scale_counts(counts, NULL, "threshold", c(1.5, 0.5)),
matrix(c(4.5, 10.5, 5.5, 1), ncol = 2, dimnames = dimnames(counts))
)
})
test_that("content_factor() floors at 1 before the geometric mean", {
rows <- matrix(c(0.5, 4, 2, 8), ncol = 2)
g <- c(exp(mean(log(c(1, 4)))), exp(mean(log(c(2, 8)))))
expect_equal(NACHO:::content_factor(rows), mean(g) / g)
})
test_that("content_factor() keeps one missing sample from spoiling the others", {
rows <- matrix(c(4, 2, NA, NA, 8, 2), ncol = 3)
factor <- NACHO:::content_factor(rows)
expect_true(is.nan(factor[[2]]))
expect_false(anyNA(factor[-2]))
})
test_that("sample_metrics() gives NA and one warning when lane attributes are missing", {
x <- GSE74821
samples <- x@samples[, !grepl("^Lane_Attributes", names(x@samples))]
warnings <- list()
metrics <- withCallingHandlers(
NACHO:::sample_metrics(x@counts, x@probes, samples, "legacy"),
nacho_warning_metric_unavailable = function(cnd) {
warnings <<- c(warnings, list(cnd))
invokeRestart("muffleWarning")
}
)
expect_length(warnings, 1)
expect_match(
conditionMessage(warnings[[1]]),
"Lane_Attributes.lane_BindingDensity",
fixed = TRUE
)
expect_named(
metrics,
c(
"Date",
"ID",
"BD",
"ScannerID",
"StagePosition",
"CartridgeID",
"FoV",
"PCL",
"LoD",
"MC",
"MedC"
)
)
expect_true(all(is.na(metrics$FoV)))
expect_true(all(is.na(metrics$BD)))
expect_false(anyNA(metrics$MC))
})
test_that("missing probes in some files keep QC working", {
fixture <- geo_fixture("GSE178516")
directory <- withr::local_tempdir()
files <- file.path(fixture$dir, fixture$samplesheet$IDFILE[1:3])
file.copy(files, directory)
target <- file.path(directory, basename(files[1]))
lines <- readLines(target)
lines <- lines[-grep("^Endogenous,", lines)[1]]
connection <- gzfile(target, "w")
writeLines(lines, connection)
close(connection)
expect_warning(
x <- suppressMessages(load_rcc(
directory,
fixture$samplesheet[1:3, ],
"IDFILE",
n_comp = 2
)),
class = "nacho_warning_missing_counts"
)
expect_identical(sum(is.na(nacho_counts(x))), 1L)
expect_true(all(is.finite(nacho_qc(x)$MC)))
expect_true(all(is.finite(nacho_qc(x)$Positive_factor)))
})
test_that("a negative probe with all counts missing is never excluded", {
exclude <- function(counts, preset) {
NACHO:::excluded_negatives(counts, rep("Negative", nrow(counts)), preset)
}
counts <- negative_matrix(c(40, 11, 9, 12, 10))
counts["NEG_E", ] <- NA
expect_false("NEG_E" %in% expect_no_error(exclude(counts, "legacy")))
few <- negative_matrix(c(40, 11, 9, 12))
few["NEG_D", ] <- NA
expect_false("NEG_D" %in% expect_no_error(exclude(few, "legacy")))
none <- negative_matrix(c(40, 11, 9, 12))
none[] <- NA
for (preset in c("nsolver", "legacy")) {
expect_identical(expect_no_error(exclude(none, preset)), character(0))
}
})
test_that("the Bruker rule floors the negative means at 1 before the ratio", {
expect_identical(exclude_nsolver(c(1, 0, 0, 0)), character(0))
expect_identical(exclude_nsolver(c(2, 1, 0, 0)), character(0))
expect_identical(exclude_nsolver(c(4, 0, 0, 0)), "NEG_A")
})
test_that("the legacy rule keeps every negative when it would drop them all", {
counts <- matrix(
c(10, 10, 100, 100),
ncol = 2,
byrow = TRUE,
dimnames = list(c("A", "B"), 1:2)
)
overall <- stats::median(counts)
medians <- apply(counts, 1, stats::median)
expect_true(all(abs(overall - medians) > 0.5 * overall))
expect_identical(
NACHO:::excluded_negatives(counts, rep("Negative", 2), "legacy"),
character(0)
)
})
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.