tests/testthat/test_function-isotope.R
In rformassspectrometry/MetaboCoreUtils: Core Utils for Metabolomics Data
library(testthat)
test_that("isotopologues works", {
substDef <- cbind(md = rep(c(1, 2), times = c(2, 3)),
leftend = c(c(50, 100), c(70, 100, 150)),
rightend = c(c(100, 150), c(100, 150, 220)),
LBint = c(0:1, 2:4),
LBslope = c(1:2, 3:5),
UBint = c(1:2, 3:5),
UBslope = c(2:3, 4:6))
bp1 <- data.frame(mz = c(80), intensity = c(80))
## construct mz's from md of above substitutions and corresponding intensities
## that should be accepted by construction
mz <- bp1$mz + c(1, 2)
intensity <- bp1$intensity*(c(0.5, 2.5) + bp1$mz * c(1.5, 3.5))
p1 <- data.frame(mz = mz, intensity = intensity)
bp2 <- data.frame(mz = c(110), intensity = c(50))
## 3rd peak doesn't correspond to any md in substDef
mz <- bp2$mz + c(1, 2, 3)
## 1st peak intensity incompatible by construction
## 2nd peak intensity compatible by construction
intensity <- bp2$intensity*(rep(3.5, 3) + bp2$mz * rep(4.5, 3))
p2 <- data.frame(mz = mz, intensity = intensity)
x <- rbind(bp1, p1, bp2, p2)
## Errors
x2 <- x[c(2, 3, 1, 6, 5, 7, 4), ]
expect_error(isotopologues(x2), "increasingly")
expect_equal(isotopologues(x, .check = FALSE), list())
x2 <- x
x2[4, 1] <- NA
expect_error(isotopologues(x2), "increasingly")
## search for all the groups
res <- isotopologues(x, substDef)
expect_equal(res, list(c(1, 2, 3), c(4, 6)))
## search for the group with mz compatible with seedMz
res <- isotopologues(x, substDef, seedMz = bp2$mz)
expect_equal(res, list(c(4,6)))
res <- isotopologues(x, substDef, seedMz = c(bp1$mz, bp2$mz))
expect_equal(res, list(c(1, 2, 3), c(4, 6)))
## ppm and tolerance
x[2, "mz"] <- x[2, "mz"] * (1 + 30 * 1e-06)
res <- isotopologues(x, substDef, seedMz = c(bp1$mz, bp2$mz))
## expect_equal(res, list(c(1, 3), c(4, 6))) this gives an error
## I think the reason is the problem of closest with duplicates = "closest"
## closest(c(81, 82), c(81.1, 82), tolerance = 0, ppm = 20, duplicates = "closest")
## returns c(NA, NA)
## closest(c(81, 82), c(81, 82), tolerance = 0, ppm = 20, duplicates = "closest")
## returns c(1, 2) correctly
res <- isotopologues(x, substDef, seedMz = c(bp1$mz, bp2$mz), ppm = 40)
expect_equal(res, list(c(1, 2, 3), c(4, 6)))
res <- isotopologues(x, substDef, seedMz = c(bp1$mz, bp2$mz), tolerance = 1e-2)
expect_equal(res, list(c(1, 2, 3), c(4, 6)))
## charge = 2
bp3 <- data.frame(mz = c(40), intensity = c(50))
## create mz compatible with the substitutions
mz <- bp3$mz + c(0.5, 1)
## as well as compatible intensities
intensity <- bp3$intensity*(c(0.5, 2.5) + bp3$mz * 2 * c(1.5, 3.5))
p3 <- data.frame(mz = mz, intensity = intensity)
x <- rbind(bp3, p3, x)
res <- isotopologues(x, substDef, charge = 2)
expect_equal(res, list(c(1, 2, 3)))
})
test_that("isotopicSubstitutionMatrix works", {
expect_error(isotopicSubstitutionMatrix("other"))
res <- isotopicSubstitutionMatrix("HMDB_NEUTRAL")
expect_true(is.data.frame(res))
})
x <- read.table(system.file("exampleSpectra/simulated_spectrum.txt",
package = "MetaboCoreUtils"))
frmls <- unique(x$frmls)
subst_def <- isotopicSubstitutionMatrix("HMDB_NEUTRAL")
test_that("isotopologues works on simulated spectra from HMDB", {
expected_groups <- lapply(frmls, function(f) which(x[, "frmls"] == f))
i_groups <- isotopologues(x[, 1:2], substDefinition = subst_def, ppm = 1)
expect_equal(i_groups, expected_groups)
})
test_that("isotopologues works on simulated spectra from HMDB + random peaks", {
set.seed(123); n = 50
noise <- data.frame(mz = runif(n, min = min(x$mz), max = max(x$mz)),
intensity = runif(n, min(x$intensity), max(x$intensity)),
frmls = rep("noise", n))
x_n <- rbind(x, noise)
x_n <- x_n[order(x_n$mz), ]
expected_groups_n <- lapply(frmls, function(f) which(x_n[,"frmls"] == f))
## low ppm
i_groups <- isotopologues(x_n[, 1:2], substDefinition = subst_def, ppm = 1)
expect_equal(i_groups, expected_groups_n) # all groups correctly identified
## higher ppm: some problems occur
i_groups <- isotopologues(x_n[, 1:2], substDefinition = subst_def, ppm = 20)
# Group 1 in additional to the expected peaks contains peaks c(20, 33, 39)
expected_groups_n[[1]]
i_groups[[1]]
sbs <- closest(x_n[c(20, 33, 39), "mz"],
x_n[i_groups[[1]][1], "mz"] + subst_def$md, ppm = 20,
tolerance = 0)
subst_def[sbs, "name"]
x_n[expected_groups_n[[1]][1], "frmls"]
# The above mentioned peaks are matched to substitutions that are not possible
# in the compound that originates the signal of the expected group.
# Group 3 coincides with expected 2nd group
expect_equal(i_groups[[3]], expected_groups_n[[2]])
# 4th found group contains all the peaks in expected group 3 except peaks
# 20, 33, 39. These peaks were mistakenly assigned to the first group found
# and therefore are no longer available in the subsequent searches
i_groups[[4]]
expected_groups_n[[3]]
# Groups 2, 5, 6, 8, 11 are composed of noise peaks whose mz differences match
# some of the substitutions
x_n[i_groups[[2]], "frmls"]
x_n[i_groups[[5]], "frmls"]
x_n[i_groups[[6]], "frmls"]
x_n[i_groups[[8]], "frmls"]
x_n[i_groups[[11]], "frmls"]
# closest(x_n[45, "mz"], subst_def$md + x_n[2, "mz"] , ppm = 20, tolerance = 0)
closest(x_n[c(53, 58), "mz"], subst_def$md + x_n[41, "mz"] , ppm = 20,
tolerance = 0)
# Group 7 coincides with expected group 4
i_groups[[7]]
expected_groups_n[[4]]
# Group 9 coincides with expected group 5
i_groups[[9]]
expected_groups_n[[5]]
# Group 10 starts with a noise peak. Unfortunately peaks 87, 93, 98, 103 match
# it and are grouped with it.
x_n[i_groups[[10]], "frmls"]
# Group 12 contains peaks of expected group 6 but it's not right because
# the monoisotopic peak (87) was mistakenly assigned to group 10 along with
# some other of its peaks
i_groups[[12]]
expected_groups_n[[6]]
})
test_that(".isotope_peaks works on second test set", {
x <- read.table(system.file("exampleSpectra",
"serine-alpha-lactose-caffeine.txt",
package = "MetaboCoreUtils"),
header = TRUE)
x <- x[order(x$mz), ]
rownames(x) <- NULL
expected_groups <- lapply(unique(x$compound),
function(f) which(x[, "compound"] == f))
subst_def <- isotopicSubstitutionMatrix("HMDB_NEUTRAL")
i_groups <- .isotope_peaks(x[, 1:2],
substDefinition = subst_def,
ppm = 10)
expect_equal(expected_groups[1], i_groups[1])
expect_equal(expected_groups[2], i_groups[2])
## closest has again problems if there are two "best matching" peaks.
expect_true(all(i_groups[[3L]] %in% expected_groups[[3L]]))
set.seed(123)
x_n <- x
x_n$mz <- x_n$mz + MsCoreUtils::ppm(x_n$mz, ppm = runif(nrow(x_n), 0, 5))
## reset the monoisotopic mz
x_n$mz[x_n$intensity == 100] <- x$mz[x_n$intensity == 100]
x_n$intensity <- x_n$intensity + rnorm(nrow(x_n), sd = 0.01)
x_n <- x_n[order(x_n$mz), ]
rownames(x_n) <- NULL
expected_groups <- lapply(unique(x_n$compound),
function(f) which(x_n[, "compound"] == f))
i_groups <- .isotope_peaks(x_n[, 1:2], ppm = 10)
expect_equal(i_groups[1], expected_groups[1])
## Not perfect for the others...
expect_true(all(i_groups[[2L]] %in% expected_groups[[2L]]))
expect_true(all(i_groups[[4L]] %in% expected_groups[[3L]]))
})
test_that(".isotope_peaks_reverse works on second test set", {
x <- read.table(system.file("exampleSpectra",
"serine-alpha-lactose-caffeine.txt",
package = "MetaboCoreUtils"),
header = TRUE)
x <- x[order(x$mz), ]
rownames(x) <- NULL
expected_groups <- lapply(unique(x$compound),
function(f) which(x[, "compound"] == f))
subst_def <- isotopicSubstitutionMatrix("HMDB_NEUTRAL")
i_groups <- .isotope_peaks_reverse(x[, 1:2],
substDefinition = subst_def,
ppm = 10)
expect_equal(expected_groups[1], i_groups[1])
expect_equal(expected_groups[2], i_groups[2])
## closest has again problems if there are two "best matching" peaks.
expect_true(all(i_groups[[3L]] %in% expected_groups[[3L]]))
set.seed(123)
x_n <- x
x_n$mz <- x_n$mz + MsCoreUtils::ppm(x_n$mz, ppm = runif(nrow(x_n), 0, 5))
## reset the monoisotopic mz
x_n$mz[x_n$intensity == 100] <- x$mz[x_n$intensity == 100]
x_n$intensity <- x_n$intensity + rnorm(nrow(x_n), sd = 0.01)
x_n <- x_n[order(x_n$mz), ]
rownames(x_n) <- NULL
expected_groups <- lapply(unique(x_n$compound),
function(f) which(x_n[, "compound"] == f))
i_groups <- .isotope_peaks_reverse(x_n[, 1:2], ppm = 10)
expect_equal(i_groups[1], expected_groups[1])
## Not perfect for the others...
expect_true(all(i_groups[[2L]] %in% expected_groups[[2L]]))
expect_true(all(i_groups[[4L]] %in% expected_groups[[3L]]))
})
test_that(".isotope_peaks_exhaustive works on second test set", {
x <- read.table(system.file("exampleSpectra",
"serine-alpha-lactose-caffeine.txt",
package = "MetaboCoreUtils"),
header = TRUE)
x <- x[order(x$mz), ]
rownames(x) <- NULL
expected_groups <- lapply(unique(x$compound),
function(f) which(x[, "compound"] == f))
subst_def <- isotopicSubstitutionMatrix("HMDB_NEUTRAL")
i_groups <- .isotope_peaks_exhaustive(x[, 1:2],
substDefinition = subst_def,
ppm = 10)
expect_equal(expected_groups[1], i_groups[1])
expect_equal(expected_groups[2], i_groups[2])
expect_equal(expected_groups[3], i_groups[3])
set.seed(123)
x_n <- x
x_n$mz <- x_n$mz + MsCoreUtils::ppm(x_n$mz, ppm = runif(nrow(x_n), 0, 5))
## reset the monoisotopic mz
x_n$mz[x_n$intensity == 100] <- x$mz[x_n$intensity == 100]
x_n$intensity <- x_n$intensity + rnorm(nrow(x_n), sd = 0.01)
x_n <- x_n[order(x_n$mz), ]
rownames(x_n) <- NULL
i_groups <- .isotope_peaks_exhaustive(x_n[, 1:2], ppm = 10)
expect_equal(i_groups[1], expected_groups[1])
expect_equal(i_groups[2], expected_groups[2])
expect_equal(i_groups[3], expected_groups[3])
})
test_that(".isotope_peaks_grouped works on second test set", {
x <- read.table(system.file("exampleSpectra",
"serine-alpha-lactose-caffeine.txt",
package = "MetaboCoreUtils"),
header = TRUE)
x <- x[order(x$mz), ]
rownames(x) <- NULL
expected_groups <- lapply(unique(x$compound),
function(f) which(x[, "compound"] == f))
subst_def <- isotopicSubstitutionMatrix("HMDB_NEUTRAL")
res <- MetaboCoreUtils:::.isotope_peaks_grouped(x[, 1:2],
substDefinition = subst_def,
ppm = 5)
expect_equal(res, expected_groups)
res2 <- MetaboCoreUtils:::.isotope_peaks_grouped(x[, 1:2],
substDefinition = subst_def,
ppm = 20)
expect_equal(res, res2)
})
test_that("availableIsotopicSubstitutionMatrix works", {
res <- availableIsotopicSubstitutionMatrix()
expect_equal(res, c("HMDB_NEGATIVE", "HMDB_NEUTRAL", "HMDB_POSITIVE"))
})
performanceTest <- function() {
library(testthat)
library(microbenchmark)
library(MetaboCoreUtils)
library(Spectra)
library(MsCoreUtils)
subst <- MetaboCoreUtils::isotopicSubstitutionMatrix("HMDB_NEUTRAL")
substm <- as.matrix(subst[, -1])
## Small example.
x <- read.table(system.file("exampleSpectra",
"serine-alpha-lactose-caffeine.txt",
package = "MetaboCoreUtils"),
header = TRUE)
x <- x[order(x$mz), ]
rownames(x) <- NULL
xm <- as.matrix(x[, 1:2])
A <- MetaboCoreUtils:::.isotope_peaks(xm, substm, ppm = 5)
B <- MetaboCoreUtils:::.isotope_peaks_exhaustive(xm, substm, ppm = 5)
C <- MetaboCoreUtils:::.isotope_peaks_reverse(xm, substm, ppm = 5)
D <- MetaboCoreUtils:::.isotope_peaks_grouped(xm, substm, ppm = 5)
expect_equal(B, D)
microbenchmark(
MetaboCoreUtils:::.isotope_peaks(xm, substm, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_exhaustive(xm, substm, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_reverse(xm, substm, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_grouped(xm, substm, ppm = 5)
)
## Unit: microseconds
## expr min lq
## MetaboCoreUtils:::.isotope_peaks(xm, substm, ppm = 5) 240.3 271.10
## MetaboCoreUtils:::.isotope_peaks_exhaustive(xm, substm, ppm = 5) 779.5 856.85
## MetaboCoreUtils:::.isotope_peaks_reverse(xm, substm, ppm = 5) 210.5 238.65
## mean median uq max neval cld
## 317.454 304.65 342.7 743.5 100 a
## 972.414 932.15 985.6 4510.1 100 b
## 275.806 276.20 296.6 397.7 100 a
sps <- Spectra("/Users/jo/Projects/git/EuracBiomedicalResearch/end-to-end-untargeted-metabolomics/data/mzML/POOL_1.mzML")
## sps <- Spectra("/data/massspec/mzML/2017/2017_04/20170403_POOL_POS_7.mzML")
x2 <- peaksData(sps[123])[[1L]]
A <- MetaboCoreUtils:::.isotope_peaks(x2, substm, ppm = 5)
B <- MetaboCoreUtils:::.isotope_peaks_exhaustive(x2, substm, ppm = 5)
C <- MetaboCoreUtils:::.isotope_peaks_reverse(x2, substm, ppm = 5)
D <- MetaboCoreUtils:::.isotope_peaks_grouped(x2, substm, ppm = 5)
microbenchmark(
MetaboCoreUtils:::.isotope_peaks(x2, substm, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_exhaustive(x2, substm, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_reverse(x2, substm, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_grouped(x2, substm, ppm = 5),
times = 10)
## Unit: milliseconds
## expr min
## MetaboCoreUtils:::.isotope_peaks(x2, substm, ppm = 5) 277.8111
## MetaboCoreUtils:::.isotope_peaks_exhaustive(x2, substm, ppm = 5) 5020.1011
## lq mean median uq max neval cld
## 282.4424 304.7626 284.9322 291.0425 479.173 10 a
## 5045.3426 5102.1704 5077.3474 5155.8181 5297.600 10 b
microbenchmark(
MetaboCoreUtils:::.isotope_peaks(x2, subst, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_exhaustive(x2, subst, ppm = 5),
times = 10)
## Unit: milliseconds
## expr min
## MetaboCoreUtils:::.isotope_peaks(x2, subst, ppm = 5) 419.2229
## MetaboCoreUtils:::.isotope_peaks_exhaustive(x2, subst, ppm = 5) 5430.3477
## lq mean median uq max neval cld
## 423.209 454.8592 435.6449 443.8982 642.5738 10 a
## 5480.699 5588.5448 5505.4043 5579.1719 6107.9860 10 b
## Summary:
## - reverse version is not working well
## - exhaustive is exact but slow.
## - grouped is faster than exhaustive, but can yield different results.
}
rformassspectrometry/MetaboCoreUtils documentation built on April 15, 2024, 10:27 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
library(testthat)
test_that("isotopologues works", {
substDef <- cbind(md = rep(c(1, 2), times = c(2, 3)),
leftend = c(c(50, 100), c(70, 100, 150)),
rightend = c(c(100, 150), c(100, 150, 220)),
LBint = c(0:1, 2:4),
LBslope = c(1:2, 3:5),
UBint = c(1:2, 3:5),
UBslope = c(2:3, 4:6))
bp1 <- data.frame(mz = c(80), intensity = c(80))
## construct mz's from md of above substitutions and corresponding intensities
## that should be accepted by construction
mz <- bp1$mz + c(1, 2)
intensity <- bp1$intensity*(c(0.5, 2.5) + bp1$mz * c(1.5, 3.5))
p1 <- data.frame(mz = mz, intensity = intensity)
bp2 <- data.frame(mz = c(110), intensity = c(50))
## 3rd peak doesn't correspond to any md in substDef
mz <- bp2$mz + c(1, 2, 3)
## 1st peak intensity incompatible by construction
## 2nd peak intensity compatible by construction
intensity <- bp2$intensity*(rep(3.5, 3) + bp2$mz * rep(4.5, 3))
p2 <- data.frame(mz = mz, intensity = intensity)
x <- rbind(bp1, p1, bp2, p2)
## Errors
x2 <- x[c(2, 3, 1, 6, 5, 7, 4), ]
expect_error(isotopologues(x2), "increasingly")
expect_equal(isotopologues(x, .check = FALSE), list())
x2 <- x
x2[4, 1] <- NA
expect_error(isotopologues(x2), "increasingly")
## search for all the groups
res <- isotopologues(x, substDef)
expect_equal(res, list(c(1, 2, 3), c(4, 6)))
## search for the group with mz compatible with seedMz
res <- isotopologues(x, substDef, seedMz = bp2$mz)
expect_equal(res, list(c(4,6)))
res <- isotopologues(x, substDef, seedMz = c(bp1$mz, bp2$mz))
expect_equal(res, list(c(1, 2, 3), c(4, 6)))
## ppm and tolerance
x[2, "mz"] <- x[2, "mz"] * (1 + 30 * 1e-06)
res <- isotopologues(x, substDef, seedMz = c(bp1$mz, bp2$mz))
## expect_equal(res, list(c(1, 3), c(4, 6))) this gives an error
## I think the reason is the problem of closest with duplicates = "closest"
## closest(c(81, 82), c(81.1, 82), tolerance = 0, ppm = 20, duplicates = "closest")
## returns c(NA, NA)
## closest(c(81, 82), c(81, 82), tolerance = 0, ppm = 20, duplicates = "closest")
## returns c(1, 2) correctly
res <- isotopologues(x, substDef, seedMz = c(bp1$mz, bp2$mz), ppm = 40)
expect_equal(res, list(c(1, 2, 3), c(4, 6)))
res <- isotopologues(x, substDef, seedMz = c(bp1$mz, bp2$mz), tolerance = 1e-2)
expect_equal(res, list(c(1, 2, 3), c(4, 6)))
## charge = 2
bp3 <- data.frame(mz = c(40), intensity = c(50))
## create mz compatible with the substitutions
mz <- bp3$mz + c(0.5, 1)
## as well as compatible intensities
intensity <- bp3$intensity*(c(0.5, 2.5) + bp3$mz * 2 * c(1.5, 3.5))
p3 <- data.frame(mz = mz, intensity = intensity)
x <- rbind(bp3, p3, x)
res <- isotopologues(x, substDef, charge = 2)
expect_equal(res, list(c(1, 2, 3)))
})
test_that("isotopicSubstitutionMatrix works", {
expect_error(isotopicSubstitutionMatrix("other"))
res <- isotopicSubstitutionMatrix("HMDB_NEUTRAL")
expect_true(is.data.frame(res))
})
x <- read.table(system.file("exampleSpectra/simulated_spectrum.txt",
package = "MetaboCoreUtils"))
frmls <- unique(x$frmls)
subst_def <- isotopicSubstitutionMatrix("HMDB_NEUTRAL")
test_that("isotopologues works on simulated spectra from HMDB", {
expected_groups <- lapply(frmls, function(f) which(x[, "frmls"] == f))
i_groups <- isotopologues(x[, 1:2], substDefinition = subst_def, ppm = 1)
expect_equal(i_groups, expected_groups)
})
test_that("isotopologues works on simulated spectra from HMDB + random peaks", {
set.seed(123); n = 50
noise <- data.frame(mz = runif(n, min = min(x$mz), max = max(x$mz)),
intensity = runif(n, min(x$intensity), max(x$intensity)),
frmls = rep("noise", n))
x_n <- rbind(x, noise)
x_n <- x_n[order(x_n$mz), ]
expected_groups_n <- lapply(frmls, function(f) which(x_n[,"frmls"] == f))
## low ppm
i_groups <- isotopologues(x_n[, 1:2], substDefinition = subst_def, ppm = 1)
expect_equal(i_groups, expected_groups_n) # all groups correctly identified
## higher ppm: some problems occur
i_groups <- isotopologues(x_n[, 1:2], substDefinition = subst_def, ppm = 20)
# Group 1 in additional to the expected peaks contains peaks c(20, 33, 39)
expected_groups_n[[1]]
i_groups[[1]]
sbs <- closest(x_n[c(20, 33, 39), "mz"],
x_n[i_groups[[1]][1], "mz"] + subst_def$md, ppm = 20,
tolerance = 0)
subst_def[sbs, "name"]
x_n[expected_groups_n[[1]][1], "frmls"]
# The above mentioned peaks are matched to substitutions that are not possible
# in the compound that originates the signal of the expected group.
# Group 3 coincides with expected 2nd group
expect_equal(i_groups[[3]], expected_groups_n[[2]])
# 4th found group contains all the peaks in expected group 3 except peaks
# 20, 33, 39. These peaks were mistakenly assigned to the first group found
# and therefore are no longer available in the subsequent searches
i_groups[[4]]
expected_groups_n[[3]]
# Groups 2, 5, 6, 8, 11 are composed of noise peaks whose mz differences match
# some of the substitutions
x_n[i_groups[[2]], "frmls"]
x_n[i_groups[[5]], "frmls"]
x_n[i_groups[[6]], "frmls"]
x_n[i_groups[[8]], "frmls"]
x_n[i_groups[[11]], "frmls"]
# closest(x_n[45, "mz"], subst_def$md + x_n[2, "mz"] , ppm = 20, tolerance = 0)
closest(x_n[c(53, 58), "mz"], subst_def$md + x_n[41, "mz"] , ppm = 20,
tolerance = 0)
# Group 7 coincides with expected group 4
i_groups[[7]]
expected_groups_n[[4]]
# Group 9 coincides with expected group 5
i_groups[[9]]
expected_groups_n[[5]]
# Group 10 starts with a noise peak. Unfortunately peaks 87, 93, 98, 103 match
# it and are grouped with it.
x_n[i_groups[[10]], "frmls"]
# Group 12 contains peaks of expected group 6 but it's not right because
# the monoisotopic peak (87) was mistakenly assigned to group 10 along with
# some other of its peaks
i_groups[[12]]
expected_groups_n[[6]]
})
test_that(".isotope_peaks works on second test set", {
x <- read.table(system.file("exampleSpectra",
"serine-alpha-lactose-caffeine.txt",
package = "MetaboCoreUtils"),
header = TRUE)
x <- x[order(x$mz), ]
rownames(x) <- NULL
expected_groups <- lapply(unique(x$compound),
function(f) which(x[, "compound"] == f))
subst_def <- isotopicSubstitutionMatrix("HMDB_NEUTRAL")
i_groups <- .isotope_peaks(x[, 1:2],
substDefinition = subst_def,
ppm = 10)
expect_equal(expected_groups[1], i_groups[1])
expect_equal(expected_groups[2], i_groups[2])
## closest has again problems if there are two "best matching" peaks.
expect_true(all(i_groups[[3L]] %in% expected_groups[[3L]]))
set.seed(123)
x_n <- x
x_n$mz <- x_n$mz + MsCoreUtils::ppm(x_n$mz, ppm = runif(nrow(x_n), 0, 5))
## reset the monoisotopic mz
x_n$mz[x_n$intensity == 100] <- x$mz[x_n$intensity == 100]
x_n$intensity <- x_n$intensity + rnorm(nrow(x_n), sd = 0.01)
x_n <- x_n[order(x_n$mz), ]
rownames(x_n) <- NULL
expected_groups <- lapply(unique(x_n$compound),
function(f) which(x_n[, "compound"] == f))
i_groups <- .isotope_peaks(x_n[, 1:2], ppm = 10)
expect_equal(i_groups[1], expected_groups[1])
## Not perfect for the others...
expect_true(all(i_groups[[2L]] %in% expected_groups[[2L]]))
expect_true(all(i_groups[[4L]] %in% expected_groups[[3L]]))
})
test_that(".isotope_peaks_reverse works on second test set", {
x <- read.table(system.file("exampleSpectra",
"serine-alpha-lactose-caffeine.txt",
package = "MetaboCoreUtils"),
header = TRUE)
x <- x[order(x$mz), ]
rownames(x) <- NULL
expected_groups <- lapply(unique(x$compound),
function(f) which(x[, "compound"] == f))
subst_def <- isotopicSubstitutionMatrix("HMDB_NEUTRAL")
i_groups <- .isotope_peaks_reverse(x[, 1:2],
substDefinition = subst_def,
ppm = 10)
expect_equal(expected_groups[1], i_groups[1])
expect_equal(expected_groups[2], i_groups[2])
## closest has again problems if there are two "best matching" peaks.
expect_true(all(i_groups[[3L]] %in% expected_groups[[3L]]))
set.seed(123)
x_n <- x
x_n$mz <- x_n$mz + MsCoreUtils::ppm(x_n$mz, ppm = runif(nrow(x_n), 0, 5))
## reset the monoisotopic mz
x_n$mz[x_n$intensity == 100] <- x$mz[x_n$intensity == 100]
x_n$intensity <- x_n$intensity + rnorm(nrow(x_n), sd = 0.01)
x_n <- x_n[order(x_n$mz), ]
rownames(x_n) <- NULL
expected_groups <- lapply(unique(x_n$compound),
function(f) which(x_n[, "compound"] == f))
i_groups <- .isotope_peaks_reverse(x_n[, 1:2], ppm = 10)
expect_equal(i_groups[1], expected_groups[1])
## Not perfect for the others...
expect_true(all(i_groups[[2L]] %in% expected_groups[[2L]]))
expect_true(all(i_groups[[4L]] %in% expected_groups[[3L]]))
})
test_that(".isotope_peaks_exhaustive works on second test set", {
x <- read.table(system.file("exampleSpectra",
"serine-alpha-lactose-caffeine.txt",
package = "MetaboCoreUtils"),
header = TRUE)
x <- x[order(x$mz), ]
rownames(x) <- NULL
expected_groups <- lapply(unique(x$compound),
function(f) which(x[, "compound"] == f))
subst_def <- isotopicSubstitutionMatrix("HMDB_NEUTRAL")
i_groups <- .isotope_peaks_exhaustive(x[, 1:2],
substDefinition = subst_def,
ppm = 10)
expect_equal(expected_groups[1], i_groups[1])
expect_equal(expected_groups[2], i_groups[2])
expect_equal(expected_groups[3], i_groups[3])
set.seed(123)
x_n <- x
x_n$mz <- x_n$mz + MsCoreUtils::ppm(x_n$mz, ppm = runif(nrow(x_n), 0, 5))
## reset the monoisotopic mz
x_n$mz[x_n$intensity == 100] <- x$mz[x_n$intensity == 100]
x_n$intensity <- x_n$intensity + rnorm(nrow(x_n), sd = 0.01)
x_n <- x_n[order(x_n$mz), ]
rownames(x_n) <- NULL
i_groups <- .isotope_peaks_exhaustive(x_n[, 1:2], ppm = 10)
expect_equal(i_groups[1], expected_groups[1])
expect_equal(i_groups[2], expected_groups[2])
expect_equal(i_groups[3], expected_groups[3])
})
test_that(".isotope_peaks_grouped works on second test set", {
x <- read.table(system.file("exampleSpectra",
"serine-alpha-lactose-caffeine.txt",
package = "MetaboCoreUtils"),
header = TRUE)
x <- x[order(x$mz), ]
rownames(x) <- NULL
expected_groups <- lapply(unique(x$compound),
function(f) which(x[, "compound"] == f))
subst_def <- isotopicSubstitutionMatrix("HMDB_NEUTRAL")
res <- MetaboCoreUtils:::.isotope_peaks_grouped(x[, 1:2],
substDefinition = subst_def,
ppm = 5)
expect_equal(res, expected_groups)
res2 <- MetaboCoreUtils:::.isotope_peaks_grouped(x[, 1:2],
substDefinition = subst_def,
ppm = 20)
expect_equal(res, res2)
})
test_that("availableIsotopicSubstitutionMatrix works", {
res <- availableIsotopicSubstitutionMatrix()
expect_equal(res, c("HMDB_NEGATIVE", "HMDB_NEUTRAL", "HMDB_POSITIVE"))
})
performanceTest <- function() {
library(testthat)
library(microbenchmark)
library(MetaboCoreUtils)
library(Spectra)
library(MsCoreUtils)
subst <- MetaboCoreUtils::isotopicSubstitutionMatrix("HMDB_NEUTRAL")
substm <- as.matrix(subst[, -1])
## Small example.
x <- read.table(system.file("exampleSpectra",
"serine-alpha-lactose-caffeine.txt",
package = "MetaboCoreUtils"),
header = TRUE)
x <- x[order(x$mz), ]
rownames(x) <- NULL
xm <- as.matrix(x[, 1:2])
A <- MetaboCoreUtils:::.isotope_peaks(xm, substm, ppm = 5)
B <- MetaboCoreUtils:::.isotope_peaks_exhaustive(xm, substm, ppm = 5)
C <- MetaboCoreUtils:::.isotope_peaks_reverse(xm, substm, ppm = 5)
D <- MetaboCoreUtils:::.isotope_peaks_grouped(xm, substm, ppm = 5)
expect_equal(B, D)
microbenchmark(
MetaboCoreUtils:::.isotope_peaks(xm, substm, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_exhaustive(xm, substm, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_reverse(xm, substm, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_grouped(xm, substm, ppm = 5)
)
## Unit: microseconds
## expr min lq
## MetaboCoreUtils:::.isotope_peaks(xm, substm, ppm = 5) 240.3 271.10
## MetaboCoreUtils:::.isotope_peaks_exhaustive(xm, substm, ppm = 5) 779.5 856.85
## MetaboCoreUtils:::.isotope_peaks_reverse(xm, substm, ppm = 5) 210.5 238.65
## mean median uq max neval cld
## 317.454 304.65 342.7 743.5 100 a
## 972.414 932.15 985.6 4510.1 100 b
## 275.806 276.20 296.6 397.7 100 a
sps <- Spectra("/Users/jo/Projects/git/EuracBiomedicalResearch/end-to-end-untargeted-metabolomics/data/mzML/POOL_1.mzML")
## sps <- Spectra("/data/massspec/mzML/2017/2017_04/20170403_POOL_POS_7.mzML")
x2 <- peaksData(sps[123])[[1L]]
A <- MetaboCoreUtils:::.isotope_peaks(x2, substm, ppm = 5)
B <- MetaboCoreUtils:::.isotope_peaks_exhaustive(x2, substm, ppm = 5)
C <- MetaboCoreUtils:::.isotope_peaks_reverse(x2, substm, ppm = 5)
D <- MetaboCoreUtils:::.isotope_peaks_grouped(x2, substm, ppm = 5)
microbenchmark(
MetaboCoreUtils:::.isotope_peaks(x2, substm, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_exhaustive(x2, substm, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_reverse(x2, substm, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_grouped(x2, substm, ppm = 5),
times = 10)
## Unit: milliseconds
## expr min
## MetaboCoreUtils:::.isotope_peaks(x2, substm, ppm = 5) 277.8111
## MetaboCoreUtils:::.isotope_peaks_exhaustive(x2, substm, ppm = 5) 5020.1011
## lq mean median uq max neval cld
## 282.4424 304.7626 284.9322 291.0425 479.173 10 a
## 5045.3426 5102.1704 5077.3474 5155.8181 5297.600 10 b
microbenchmark(
MetaboCoreUtils:::.isotope_peaks(x2, subst, ppm = 5),
MetaboCoreUtils:::.isotope_peaks_exhaustive(x2, subst, ppm = 5),
times = 10)
## Unit: milliseconds
## expr min
## MetaboCoreUtils:::.isotope_peaks(x2, subst, ppm = 5) 419.2229
## MetaboCoreUtils:::.isotope_peaks_exhaustive(x2, subst, ppm = 5) 5430.3477
## lq mean median uq max neval cld
## 423.209 454.8592 435.6449 443.8982 642.5738 10 a
## 5480.699 5588.5448 5505.4043 5579.1719 6107.9860 10 b
## Summary:
## - reverse version is not working well
## - exhaustive is exact but slow.
## - grouped is faster than exhaustive, but can yield different results.
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.