Representation of chromatographic MS data

Description

The Chromatogram class is designed to store chromatographic MS data, i.e. pairs of retention time and intensity values. Instances of the class can be created with the Chromatogram constructor function but in most cases the dedicated methods for OnDiskMSnExp and MSnExp objects extracting chromatograms should be used instead (i.e. the chromatogram method).

Chromatogram : create an instance of the Chromatogram class.

aggregationFun,aggregationFun<- get or set the aggregation function.

rtime returns the retention times for the rentention time

• intensity pairs stored in the chromatogram.

intensity returns the intensity for the rentention time

• intensity pairs stored in the chromatogram.

mz get the mz (range) of the chromatogram. The function returns a numeric(2) with the lower and upper mz value.

precursorMz get the mz of the precursor ion. The function returns a numeric(2) with the lower and upper mz value.

fromFile returns the value from the fromFile slot.

length returns the length (number of retention time - intensity pairs) of the chromatogram.

as.data.frame returns the rtime and intensity values from the object as data.frame .

filterRt : filters the chromatogram based on the provided retention time range.

clean : Removes unused 0-intensity data points. See clean documentation for more details and examples.

plot : plots a Chromatogram object.

msLevel returns the MS level of the chromatogram.

isEmpty returns TRUE for empty chromatogram or chromatograms with all intensities being NA .

productMz get the mz of the product chromatogram/ion. The function returns a numeric(2) with the lower and upper mz value.

bin aggregates intensity values from a chromatogram in discrete bins along the retention time axis and returns a Chromatogram object with the retention time representing the mid-point of the bins and the intensity the binned signal.

Usage

Chromatogram(rtime = numeric(), intensity = numeric(),
  mz = c(NA_real_, NA_real_), filterMz = c(NA_real_, NA_real_),
  precursorMz = c(NA_real_, NA_real_), productMz = c(NA_real_,
  NA_real_), fromFile = integer(), aggregationFun = character(),
  msLevel = 1L)
aggregationFun(object)
list(list("show"), list("Chromatogram"))(object)
list(list("rtime"), list("Chromatogram"))(object)
list(list("intensity"), list("Chromatogram"))(object)
list(list("mz"), list("Chromatogram"))(object, filter = FALSE)
list(list("precursorMz"), list("Chromatogram"))(object)
list(list("fromFile"), list("Chromatogram"))(object)
list(list("length"), list("Chromatogram"))(x)
list(list("as.data.frame"), list("Chromatogram"))(x)
list(list("filterRt"), list("Chromatogram"))(object, rt)
list(list("clean"), list("Chromatogram"))(object, all = FALSE, na.rm = FALSE)
list(list("plot"), list("Chromatogram,ANY"))(x, col = "#00000060", lty = 1,
  type = "l", xlab = "retention time", ylab = "intensity",
  main = NULL, ...)
list(list("msLevel"), list("Chromatogram"))(object)
list(list("isEmpty"), list("Chromatogram"))(x)
list(list("productMz"), list("Chromatogram"))(object)
list(list("bin"), list("Chromatogram"))(object, binSize = 0.5,
  breaks = seq(floor(min(rtime(object))), ceiling(max(rtime(object))), by
  = binSize), fun = max)

Arguments

Details

The mz , filterMz , precursorMz and productMz are stored as a numeric(2) representing a range even if the chromatogram was generated for only a single ion (i.e. a single mz value). Using ranges for mz values allow this class to be used also for e.g. total ion chromatograms or base peak chromatograms.

The slots precursorMz and productMz allow to represent SRM (single reaction monitoring) and MRM (multiple SRM) chromatograms. As example, a Chromatogram for a SRM transition 273 -> 153 will have a @precursorMz = c(273, 273) and a @productMz = c(153, 153) .

Seealso

Chromatograms for combining Chromatogram in a two-dimensional matrix (rows being mz-rt ranges, columns samples). chromatogram for the method to extract chromatogram data from a MSnExp or OnDiskMSnExp object. clean for the method to clean a Chromatogram object.

Author

Johannes Rainer

Examples

## Create a simple Chromatogram object.
ints <- abs(rnorm(100, sd = 100))
rts <- seq_len(length(ints))
chr <- Chromatogram(rtime = rts, intensity = ints)
chr

## Extract intensities
intensity(chr)

## Extract retention times
rtime(chr)

## Extract the mz range - is NA for the present example
mz(chr)

## plot the Chromatogram
plot(chr)

## Create a simple Chromatogram object based on random values.
chr <- Chromatogram(intensity = abs(rnorm(1000, mean = 2000, sd = 200)),
rtime = sort(abs(rnorm(1000, mean = 10, sd = 5))))
chr

## Get the intensities
head(intensity(chr))

## Get the retention time
head(rtime(chr))

## What is the retention time range of the object?
range(rtime(chr))

## Filter the chromatogram to keep only values between 4 and 10 seconds
chr2 <- filterRt(chr, rt = c(4, 10))

range(rtime(chr2))

Container for multiple Chromatogram objects

Description

The Chromatograms class allows to store Chromatogram objects in a matrix -like two-dimensional structure.

Chromatograms : create an instance of class Chromatograms .

Chromatograms objects can, just like a matrix , be subsetted using the [ method. Single elements, rows or columns can be replaced using e.g. x[1, 1] <- value where value has to be a Chromatogram object or a list of such objects.

plot : plots a Chromatograms object. For each row in the object one plot is created, i.e. all Chromatogram objects in the same row are added to the same plot.

phenoData : accesses the phenotypical desccription of the samples. Returns an AnnotatedDataFrame object.

pData : accesses the phenotypical description of the samples. Returns a data.frame .

pData<- : replace the phenotype data.

$ and $<- : get or replace individual columns of the object's pheno data.

colnames<- : replace or set the column names of the Chromatograms object. Does also set the rownames of the phenoData .

sampleNames : get the sample names.

sampleNames<- : replace or set the sample names of the Chromatograms object (i.e. the rownames of the pheno data and colnames of the data matrix.

isEmpty : returns TRUE if the Chromatograms object or all of its Chromatogram objects is/are empty or contain only NA intensities.

featureNames : returns the feature names of the Chromatograms object.

featureNames<- : set the feature names.

featureData : return the feature data.

featureData<- : replace the object's feature data.

fData : return the feature data as a data.frame .

fData<- : replace the object's feature data by passing a data.frame

fvarLabels : return the feature data variable names (i.e. column names).

rownames<- : replace the rownames (and featureNames) of the object.

precursorMz : return the precursor m/z from the chromatograms. The method returns a matrix with 2 columns ( "mzmin" and "mzmax" ) and as many rows as there are rows in the Chromatograms object. Each row contains the precursor m/z of the chromatograms in that row. An error is thrown if the chromatograms within one row have different precursor m/z values.

productMz : return the product m/z from the chromatograms. The method returns a matrix with 2 columns ( "mzmin" and "mzmax" ) and as many rows as there are rows in the Chromatograms object. Each row contains the product m/z of the chromatograms in that row. An error is thrown if the chromatograms within one row have different product m/z values.

mz : returns the m/z for each row of the Chromatograms object as a two-column matrix (with columns "mzmin" and "mzmax" ).

polarity : returns the polarity of the scans/chromatograms: 1, 0 or -1 for positive, negative or unknown polarity.

bin aggregates intensity values of chromatograms in discrete bins along the retention time axis. By default, individual Chromatogram objects of one row are binned into the same bins. The function returns a Chromatograms object with binned chromatograms.

clean : removes 0-intensity data points. Either all of them (with all = TRUE ) or all except those adjacent to non-zero intensities ( all = FALSE ; default). See clean documentation for more details and examples.

Usage

Chromatograms(data, phenoData, featureData, ...)
list(list("show"), list("Chromatograms"))(object)
list(list("["), list("Chromatograms,ANY,ANY,ANY"))(x, i, j, drop = FALSE)
list(list("["), list("Chromatograms"))(x, i, j) <- value
list(list("plot"), list("Chromatograms,ANY"))(x, col = "#00000060", lty = 1,
  type = "l", xlab = "retention time", ylab = "intensity",
  main = NULL, ...)
list(list("phenoData"), list("Chromatograms"))(object)
list(list("pData"), list("Chromatograms"))(object)
list(list("pData"), list("Chromatograms,data.frame"))(object) <- value
list(list("$"), list("Chromatograms"))(x, name)
list(list("$"), list("Chromatograms"))(x, name) <- value
list(list("colnames"), list("Chromatograms"))(x) <- value
list(list("sampleNames"), list("Chromatograms"))(object)
list(list("sampleNames"), list("Chromatograms,ANY"))(object) <- value
list(list("isEmpty"), list("Chromatograms"))(x)
list(list("featureNames"), list("Chromatograms"))(object)
list(list("featureNames"), list("Chromatograms"))(object) <- value
list(list("featureData"), list("Chromatograms"))(object)
list(list("featureData"), list("Chromatograms,ANY"))(object) <- value
list(list("fData"), list("Chromatograms"))(object)
list(list("fData"), list("Chromatograms,ANY"))(object) <- value
list(list("fvarLabels"), list("Chromatograms"))(object)
list(list("rownames"), list("Chromatograms"))(x) <- value
list(list("precursorMz"), list("Chromatograms"))(object)
list(list("productMz"), list("Chromatograms"))(object)
list(list("mz"), list("Chromatograms"))(object)
list(list("polarity"), list("Chromatograms"))(object)
list(list("bin"), list("Chromatograms"))(object, binSize = 0.5,
  breaks = numeric(), fun = max)
list(list("clean"), list("Chromatograms"))(object, all = FALSE, na.rm = FALSE)

Arguments

Details

The Chromatograms class extends the base matrix class and hence allows to store Chromatogram objects in a two-dimensional array. Each row is supposed to contain Chromatogram objects for one MS data slice with a common m/z and rt range. Columns contain Chromatogram objects from the same sample.

plot : if nrow(x) > 1 the plot area is split into nrow(x) sub-plots and the chromatograms of one row are plotted in each.

Value

For [ : the subset of the Chromatograms object. If a single element is extracted (e.g. if i and j are of length 1) a Chromatogram object is returned. Otherwise (if drop = FALSE , the default, is specified) a Chromatograms object is returned. If drop = TRUE is specified, the method returns a list of Chromatogram objects.

For phenoData : an AnnotatedDataFrame representing the pheno data of the object.

For pData : a data.frame representing the pheno data of the object.

For $ : the value of the corresponding column in the pheno data table of the object.

Seealso

Chromatogram for the class representing chromatogram data. chromatogram for the method to extract a Chromatograms object from a MSnExp or OnDiskMSnExp object. readSRMData for the function to read chromatographic data of an SRM/MRM experiment.

Note

Subsetting with [ will always return a Chromatograms object (with the exception of extracting a single element) unless drop = TRUE is specified. This is different from the default subsetting behaviour of matrix -like objects.

Author

Johannes Rainer

Examples

## Creating some chromatogram objects to put them into a Chromatograms object
ints <- abs(rnorm(25, sd = 200))
ch1 <- Chromatogram(rtime = 1:length(ints), ints)
ints <- abs(rnorm(32, sd = 90))
ch2 <- Chromatogram(rtime = 1:length(ints), ints)
ints <- abs(rnorm(19, sd = 120))
ch3 <- Chromatogram(rtime = 1:length(ints), ints)
ints <- abs(rnorm(21, sd = 40))
ch4 <- Chromatogram(rtime = 1:length(ints), ints)

## Create a Chromatograms object with 2 rows and 2 columns
chrs <- Chromatograms(list(ch1, ch2, ch3, ch4), nrow = 2)
chrs

## Extract the first element from the second column. Extracting a single
## element always returns a Chromatogram object.
chrs[1, 2]

## Extract the second row. Extracting a row or column (i.e. multiple elements
## returns by default a list of Chromatogram objects.
chrs[2, ]

## Extract the second row with drop = FALSE, i.e. return a Chromatograms
## object.
chrs[2, , drop = FALSE]

## Replace the first element.
chrs[1, 1] <- ch3
chrs

## Add a pheno data.
pd <- data.frame(name = c("first sample", "second sample"),
idx = 1:2)
pData(chrs) <- pd

## Column names correspond to the row names of the pheno data
chrs

## Access a column within the pheno data
chrs$name

## Access the m/z ratio for each row; this will be NA for the present
## object
mz(chrs)

## Create some random Chromatogram objects
ints <- abs(rnorm(123, mean = 200, sd = 32))
ch1 <- Chromatogram(rtime = seq_along(ints), intensity = ints, mz = 231)
ints <- abs(rnorm(122, mean = 250, sd = 43))
ch2 <- Chromatogram(rtime = seq_along(ints), intensity = ints, mz = 231)
ints <- abs(rnorm(125, mean = 590, sd = 120))
ch3 <- Chromatogram(rtime = seq_along(ints), intensity = ints, mz = 542)
ints <- abs(rnorm(124, mean = 1200, sd = 509))
ch4 <- Chromatogram(rtime = seq_along(ints), intensity = ints, mz = 542)

## Combine into a 2x2 Chromatograms object
chrs <- Chromatograms(list(ch1, ch2, ch3, ch4), byrow = TRUE, ncol = 2)

## Plot the second row
plot(chrs[2, , drop = FALSE])

## Plot all chromatograms
plot(chrs, col = c("#ff000080", "#00ff0080"))

Class "FeatComp"

Description

Comparing feature names of two comparable MSnSet instances.

Seealso

averageMSnSet to compuate an average MSnSet .

Author

Laurent Gatto lg390@cam.ac.uk and Thomas Naake

Examples

library("pRolocdata")
data(tan2009r1)
data(tan2009r2)
x <- compfnames(tan2009r1, tan2009r2)
x[[1]]
x[2:3]
head(common(x[[1]]))

data(tan2009r3)
tanl <- list(tan2009r1, tan2009r2, tan2009r3)
xx <- compfnames(tanl, fcol1 = NULL)
length(xx)
tail(xx)

all.equal(xx[[15]],
compfnames(tan2009r2, tan2009r3, fcol1 = NULL))
str(sapply(xx, common))

Features of Interest

Description

The Features of Interest infrastructure allows to define a set of features of particular interest to be used/matched against existing data sets contained in " . A specific set of features is stored as an FeaturesOfInterest object and a collection of such non-redundant instances (for example for a specific organism, project, ...) can be collected in a FoICollection .

Author

Laurent Gatto lg390@cam.ac.uk

Examples

library("pRolocdata")
data(tan2009r1)

x <- FeaturesOfInterest(description = "A traceable test set of features of interest",
fnames = featureNames(tan2009r1)[1:10],
object = tan2009r1)
x

description(x)
foi(x)

y <- FeaturesOfInterest(description = "Non-traceable features of interest",
fnames = featureNames(tan2009r1)[111:113])
y

## an illegal FeaturesOfInterest
try(FeaturesOfInterest(description = "Won't work",
fnames = c("A", "Z", featureNames(tan2009r1)),
object = tan2009r1))


FeaturesOfInterest(description = "This work, but not traceable",
fnames = c("A", "Z", featureNames(tan2009r1)))



xx <- FoICollection()
xx

xx <- addFeaturesOfInterest(x, xx)
xx <- addFeaturesOfInterest(y, xx)
names(xx) <- LETTERS[1:2]
xx

## Sub-setting
xx[1]
xx[[1]]
xx[["A"]]

description(xx)
foi(xx)

fnamesIn(x, tan2009r1)
fnamesIn(x, tan2009r1, count = TRUE)

rmFeaturesOfInterest(xx, 1)

The "MIAPE" Class for Storing Proteomics Experiment Information

Description

The Minimum Information About a Proteomics Experiment. The current implementation is based on the MIAPE-MS 2.4 document.

Author

Laurent Gatto lg390@cam.ac.uk

References

About MIAPE: http://www.psidev.info/index.php?q=node/91 , and references therein, especially 'Guidelines for reporting the use of mass spectrometry in proteomics', Nature Biotechnology 26, 860-861 (2008).

Class MSmap

Description

A class to store mass spectrometry data maps, i.e intensities collected along the M/Z and retention time space during a mass spectrometry acquisition.

Author

Laurent Gatto lg390@cam.ac.uk

Examples

## downloads the data
library("rpx")
px1 <- PXDataset("PXD000001")
(i <- grep("TMT.+mzML", pxfiles(px1), value = TRUE))
mzf <- pxget(px1, i)

## Using an mzRpwiz object
## reads the data
ms <- openMSfile(mzf)
hd <- header(ms)

## a set of spectra of interest: MS1 spectra eluted
## between 30 and 35 minutes retention time
ms1 <- which(hd$msLevel == 1)
rtsel <- hd$retentionTime[ms1] / 60 > 30 &
hd$retentionTime[ms1] / 60 < 35

## the map
M <- MSmap(ms, ms1[rtsel], 521, 523, .005, hd)

plot(M, aspect = 1, allTicks = FALSE)
plot3D(M)
if (require("rgl") & interactive())
plot3D(M, rgl = TRUE)

## With some MS2 spectra
i <- ms1[which(rtsel)][1]
j <- ms1[which(rtsel)][2]
M2 <- MSmap(ms, i:j, 100, 1000, 1, hd)
plot3D(M2)

## Using an OnDiskMSnExp object and accessors
msn <- readMSData(mzf, mode = "onDisk")

## a set of spectra of interest: MS1 spectra eluted
## between 30 and 35 minutes retention time
ms1 <- which(msLevel(msn) == 1)
rtsel <- rtime(msn)[ms1] / 60 > 30 &
rtime(msn)[ms1] / 60 < 35

## the map
M3 <- MSmap(msn, ms1[rtsel], 521, 523, .005)
plot(M3, aspect = 1, allTicks = FALSE)

## With some MS2 spectra
i <- ms1[which(rtsel)][1]
j <- ms1[which(rtsel)][2]
M4 <- MSmap(msn, i:j, 100, 1000, 1)
plot3D(M4)

The 'MSnExp' Class for MS Data And Meta-Data

Description

The MSnExp class encapsulates data and meta-data for mass spectrometry experiments, as described in the slots section. Several data files (currently in mzXML ) can be loaded together with the function readMSData .

This class extends the virtual " class.

In version 1.19.12, the polarity slot had been added to the " class (previously in " ). Hence, "MSnExp" objects created prior to this change will not be valid anymore, since all MS2 spectra will be missing the polarity slot. Object can be appropriately updated using the updateObject method.

The feature variables in the feature data slot will depend on the file. See also the documentation in the mzR package that parses the raw data files and produces these data.

Seealso

" and readMSData for loading mzXML , mzData or mzML files to generate an instance of MSnExp .

The " manual page contains further details and examples.

chromatogram to extract chromatographic data from a MSnExp or OnDiskMSnExp object.

write for the function to write the data to mzML or mzXML file(s).

Author

Laurent Gatto lg390@cam.ac.uk

References

Information about the mzXML format as well converters from vendor specific formats to mzXML: http://tools.proteomecenter.org/wiki/index.php?title=Formats:mzXML .

Examples

mzxmlfile <- dir(system.file("extdata",package="MSnbase"),
pattern="mzXML",full.names=TRUE)
msnexp <- readMSData(mzxmlfile)
msnexp

The "MSnProcess" Class

Description

MSnProcess is a container for MSnExp and MSnSet processing information. It records data files, processing steps, thresholds, analysis methods and times that have been applied to MSnExp or MSnSet instances.

Seealso

See the " and " classes that actually use MSnProcess as a slot.

Note

This class is likely to be updated using an AnnotatedDataFrame .

Author

Laurent Gatto lg390@cam.ac.uk

Examples

showClass("MSnProcess")

Storing multiple related MSnSets

Description

A class for storing lists of MSnSet instances.

Details

There are two ways to store different sets of measurements pertaining an experimental unit, such as replicated measures of different conditions that were recorded over more than one MS acquisition. Without focusing on any proteomics technology in particular, these multiple assays can be recorded as

• A single combined MSnSet (see the section Combining MSnSet instances in the MSnbase-demo section). In such cases, the different experimental (phenotypical) conditions are recorded as an AnnotatedDataFrame in the phenoData slots. Quantitative data for features that were missing in an assay are generally encode as missing with NA values. Alternatively, only features observed in all assays could be selected. See the commonFeatureNames functions to select only common features among two or more MSnSet instance.

• Each set of measurements is stored in an MSnSet which are combined into one MSnSetList . Each MSnSet elements can have identical or different samples and features. Unless compiled directly manually by the user, one would expect at least one of these dimensions (features/rows or samples/columns) are conserved (i.e. all feature or samples names are identical). See split / unsplit below.

Seealso

The commonFeatureNames function to select common features among MSnSet instances.

Author

Laurent Gatto lg390@cam.ac.uk

Examples

library("pRolocdata")
data(tan2009r1)
data(tan2009r2)

## The MSnSetList class
##  for an unnamed list, names are set to indices
msnl <- MSnSetList(list(tan2009r1, tan2009r2))
names(msnl)
##  a named example
msnl <- MSnSetList(list(A = tan2009r1, B = tan2009r2))
names(msnl)
msnsets(msnl)
length(msnl)
objlog(msnl)
msnl[[1]] ## an MSnSet
msnl[1]   ## an MSnSetList of length 1

## Iterating over the elements
lapply(msnl, dim) ## a list
lapply(msnl, normalise) ## an MSnSetList

fData(msnl)
fData(msnl)$X <- sapply(msnl, nrow)
fData(msnl)

## Splitting and unsplitting
##  splitting along the columns/samples
data(dunkley2006)
head(pData(dunkley2006))
(splt <- split(dunkley2006, "replicate"))
lapply(splt, dim) ## the number of rows and columns of the split elements
unsplt <- unsplit(splt, dunkley2006$replicate)
stopifnot(compareMSnSets(dunkley2006, unsplt))

##  splitting along the rows/features
head(fData(dunkley2006))
(splt <- split(dunkley2006, "markers"))
unsplt <- unsplit(splt, factor(fData(dunkley2006)$markers))
simplify2array(lapply(splt, dim))
stopifnot(compareMSnSets(dunkley2006, unsplt))

The "MSnSet" Class for MS Proteomics Expression Data and Meta-Data

Description

The MSnSet holds quantified expression data for MS proteomics data and the experimental meta-data. The MSnSet class is derived from the " class and mimics the " class classically used for microarray data.

Seealso

" , " and quantify . MSnSet quantitation values and annotation can be exported to a file with write.exprs . See readMSnSet to create and MSnSet using data available in a spreadsheet or data.frame .

Author

Laurent Gatto lg390@cam.ac.uk

Examples

data(msnset)
msnset <- msnset[10:15]

exprs(msnset)[1, c(1, 4)] <- NA
exprs(msnset)[2, c(1, 2)] <- NA
is.na(msnset)
featureNames(filterNA(msnset, pNA = 1/4))
featureNames(filterNA(msnset, pattern = "0110"))

M <- matrix(rnorm(12), 4)
pd <- data.frame(otherpdata = letters[1:3])
fd <- data.frame(otherfdata = letters[1:4])
x0 <- MSnSet(M, fd, pd)
sampleNames(x0)

M <- matrix(rnorm(12), 4)
colnames(M) <- LETTERS[1:3]
rownames(M) <- paste0("id", LETTERS[1:4])
pd <- data.frame(otherpdata = letters[1:3])
rownames(pd) <- colnames(M)
fd <- data.frame(otherfdata = letters[1:4])
rownames(fd) <- rownames(M)
x <- MSnSet(M, fd, pd)
sampleNames(x)


## Visualisation

library("pRolocdata")
data(dunkley2006)
image(dunkley2006)
## Changing colours
image(dunkley2006, high = "darkgreen")
image(dunkley2006, high = "darkgreen", low = "yellow")
## Forcing feature names
image(dunkley2006, fnames = TRUE)
## Facetting
image(dunkley2006, facetBy = "replicate")
p <- image(dunkley2006)
library("ggplot2") ## for facet_grid
p + facet_grid(replicate ~ membrane.prep, scales = 'free', space = 'free')
p + facet_grid(markers ~ replicate)
## Fold-changes
dd <- dunkley2006
exprs(dd) <- exprs(dd) - 0.25
image(dd)
image(dd, low = "green", high = "red")
## Feature names are displayed by default for smaller data
dunkley2006 <- dunkley2006[1:25, ]
image(dunkley2006)
image(dunkley2006, legend = "hello")


## Coercion
if (require("SummarizedExperiment")) {
data(msnset)
se <- as(msnset, "SummarizedExperiment")
metadata(se) ## only logging
se <- addMSnSetMetadata(se, msnset)
metadata(se) ## all metadata
msnset2 <- as(se, "MSnSet")
processingData(msnset2)
}

as(msnset, "ExpressionSet")

MSnbase options

Description

MSnbase defined a few options globally using the standard R options mechanism. The current values of these options can be queried with MSnbaseOptions . The options are:

• verbose : defines a session-wide verbosity flag, that is used if the verbose argument in individual functions is not set.

• PARALLEL_THRESH : defines the minimum number of spectra per file necessary before using parallel processing.

• fastLoad : logical(1) . If TRUE performs faster data loading for all methods of OnDiskMSnExp that load data from the original files (such as spectrapply() ). Users experiencing data I/O errors (observed mostly on macOS systems) should set this option to FALSE .

Usage

MSnbaseOptions()
isMSnbaseVerbose()
setMSnbaseVerbose(opt)
setMSnbaseParallelThresh(opt = 1000)
setMSnbaseFastLoad(opt = TRUE)
isMSnbaseFastLoad()

Arguments

Details

isMSnbaseVerbose is one wrapper for the verbosity flag, also available through options("MSnbase")$verbose .

There are also setters to set options individually. When run without argument, the verbosity setter inverts the current value of the option.

Value

A list of MSnbase options and the single option values for the individual accessors.

Parse MzTab files

Description

The MzTab class stores the output of a basic parsing of a mzTab file. It contain the metadata (a list ), comments (a character vector), and the at least of of the following data types: proteins, peptides, PSMs and small molecules (as data.frames ).

At this stage, the metadata and data are only minimally parsed. More specific data extraction and preparation are delegated to more specialised functions, such as the as(., to = "MSnSetList") and readMzTabData for proteomics data.

Note that no attempts are made to verify the validitiy of the mzTab file.

Author

Laurent Gatto, with contributions from Richard Cotton (see https://github.com/lgatto/MSnbase/issues/41 ).

References

The mzTab format is a light-weight, tab-delimited file format for proteomics data. See https://github.com/HUPO-PSI/mzTab for details and specifications.

Griss J, Jones AR, Sachsenberg T, Walzer M, Gatto L, Hartler J, Thallinger GG, Salek RM, Steinbeck C, Neuhauser N, Cox J, Neumann S, Fan J, Reisinger F, Xu QW, Del Toro N, Perez-Riverol Y, Ghali F, Bandeira N, Xenarios I, Kohlbacher O, Vizcaino JA, Hermjakob H. The mzTab data exchange format: communicating mass-spectrometry-based proteomics and metabolomics experimental results to a wider audience. Mol Cell Proteomics. 2014 Oct;13(10):2765-75. doi: 10.1074/mcp.O113.036681. Epub 2014 Jun 30. PubMed PMID: 24980485; PubMed Central PMCID: PMC4189001.

Examples

## Test files from the mzTab developement repository
fls <- c("Cytidine.mzTab", "MTBLS2.mztab",
"PRIDE_Exp_Complete_Ac_1643.xml-mztab.txt",
"PRIDE_Exp_Complete_Ac_16649.xml-mztab.txt",
"SILAC_CQI.mzTab", "SILAC_SQ.mzTab",
"iTRAQ_CQI.mzTab", "iTRAQ_SQI.mzTab",
"labelfree_CQI.mzTab", "labelfree_SQI.mzTab",
"lipidomics-HFD-LD-study-PL-DG-SM.mzTab",
"lipidomics-HFD-LD-study-TG.mzTab")

baseUrl <- "https://raw.githubusercontent.com/HUPO-PSI/mzTab/master/examples/1_0-Proteomics-Release/"

## a list of mzTab objects
mzt <- sapply(file.path(baseUrl, fls), MzTab)
stopifnot(length(mzt) == length(fls))
mzt[[4]]

dim(proteins(mzt[[4]]))
dim(psms(mzt[[4]]))

prots4 <- proteins(mzt[[4]])
class(prots4)
prots4[1:5, 1:4]

The OnDiskMSnExp Class for MS Data And Meta-Data

Description

Like the MSnExp class, the OnDiskMSnExp class encapsulates data and meta-data for mass spectrometry experiments, but does, in contrast to the former, not keep the spectrum data in memory, but fetches the M/Z and intensity values on demand from the raw files. This results in some instances to a reduced performance, has however the advantage of a much smaller memory footprint.

Details

The OnDiskMSnExp object stores many spectrum related information into the featureData , thus, some calls, like rtime to retrieve the retention time of the individual scans does not require the raw data to be read. Only M/Z and intensity values are loaded on-the-fly from the original files. Extraction of values for individual scans is, for mzML files, very fast. Extraction of the full data (all spectra) are performed in a per-file parallel processing strategy.

Data manipulations related to spectras' M/Z or intensity values (e.g. removePeaks or clean ) are (for OnDiskMSnExp objects) not applied immediately, but are stored for later execution into the spectraProcessingQueue . The manipulations are performed on-the-fly upon data retrieval. Other manipulations, like removal of individual spectra are applied directly, since the corresponding data is available in the object's featureData slot.

Seealso

pSet , MSnExp , readMSData

Author

Johannes Rainer johannes.rainer@eurac.edu

Examples

## Get some example mzML files
library(msdata)
mzfiles <- c(system.file("microtofq/MM14.mzML", package="msdata"),
system.file("microtofq/MM8.mzML", package="msdata"))
## Read the data as an OnDiskMSnExp
odmse <- readMSData(mzfiles, msLevel=1, centroided = TRUE)

## Get the length of data, i.e. the total number of spectra.
length(odmse)

## Get the MS level
head(msLevel(odmse))

## Get the featureData, use fData to return as a data.frame
head(fData(odmse))

## Get to know from which file the spectra are
head(fromFile(odmse))

## And the file names:
fileNames(odmse)

## Scan index and acquisitionNum
head(scanIndex(odmse))
head(acquisitionNum(odmse))

## Extract the spectra; the data is retrieved from the raw files.
head(spectra(odmse))

## Extracting individual spectra or a subset is much faster.
spectra(odmse[1:50])

## Alternatively, we could also subset the whole object by spectra and/or samples:
subs <- odmse[rtime(odmse) >= 2 & rtime(odmse) <= 20, ]
fileNames(subs)
rtime(subs)

## Extract intensities and M/Z values per spectrum; the methods return a list,
## each element representing the values for one spectrum.
ints <- intensity(odmse)
mzs <- mz(odmse)

## Return a data.frame with mz and intensity pairs for each spectrum from the
## object
res <- spectrapply(odmse, FUN = as, Class = "data.frame")

## Calling removePeaks, i.e. setting intensity values below a certain threshold to 0.
## Unlike the name suggests, this is not actually removing peaks. Such peaks with a 0
## intensity are then removed by the "clean" step.
## Also, the manipulations are not applied directly, but put into the "lazy"
## processing queue.
odmse <- removePeaks(odmse, t=10000)
odmse <- clean(odmse)

## The processing steps are only applied when actual raw data is extracted.
spectra(odmse[1:2])

## Get the polarity of the spectra.
head(polarity(odmse))

## Get the retention time of all spectra
head(rtime(odmse))

## Get the intensities after removePeaks and clean
intsAfter <- intensity(odmse)

head(lengths(ints))
head(lengths(intsAfter))

## The same for the M/Z values
mzsAfter <- intensity(odmse)
head(lengths(mzs))
head(lengths(mzsAfter))


## Centroided or profile mode
f <- msdata::proteomics(full.names = TRUE,
pattern = "MS3TMT11.mzML")
odmse <- readMSData(f, mode = "onDisk")
validObject(odmse)
odmse[[1]]

table(isCentroidedFromFile(odmse), msLevel(odmse))

## centroided status could be set manually
centroided(odmse, msLevel = 1) <- FALSE
centroided(odmse, msLevel = 2) <- TRUE
centroided(odmse, msLevel = 3) <- TRUE

## or when reading the data
odmse2 <- readMSData(f, centroided = c(FALSE, TRUE, TRUE),
mode = "onDisk")
table(centroided(odmse), msLevel(odmse))

## Filtering precursor scans

head(acquisitionNum(odmse))
head(msLevel(odmse))

## Extract all spectra stemming from the first MS1 spectrum
(from1 <- filterPrecursorScan(odmse, 21945))
table(msLevel(from1))


## Extract the second sepctrum's parent (MS1) and children (MS3)
## spectra
(from2 <- filterPrecursorScan(odmse, 21946))
table(msLevel(from2))

Simple processing step class

Description

The ProcessingStep class is a simple object to encapsule all relevant information of a data analysis processing step, i.e. the function name and all arguments.

Details

Objects of this class are mainly used to record all possible processing steps of an OnDiskMSnExp object for later lazy execution .

Seealso

OnDiskMSnExp

Author

Johannes Rainer johannes.rainer@eurac.edu

Examples

## Define a simple ProcessingStep
procS <- ProcessingStep("sum", list(c(1, 3, NA, 5), na.rm= TRUE))

executeProcessingStep(procS)

The "ReporterIons" Class

Description

The ReporterIons class allows to define a set of isobaric reporter ions that are used for quantification in MSMS mode, e.g. iTRAQ (isobaric tag for relative and absolute quantitation) or TMT (tandem mass tags). ReporterIons instances can them be used when quantifying " data of plotting the reporters peaks based on in " ojects.

Some reporter ions are provided with MSnbase an can be loaded with the data function. These reporter ions data sets are: list(" ", " ", list(list(list("iTRAQ4"), ":"), list(list("ReporterIon"), " object for the iTRAQ ", " 4-plex set. Load with ", list("data(iTRAQ4)"), ". ")), " ", " ", list(list(list("iTRAQ5"), ":"), list(list("ReporterIon"), " object for the iTRAQ ", " 4-plex set plus the isobaric tag. Load with ", list("data(iTRAQ5)"), ". ")), " ", " ", list(list(list("TMT6"), ":"), list(list("ReporterIon"), " object for the TMT ", " 6-plex set. Load with ", list("data(TMT6)"), ". ")),

"

", " ", list(list(list("TMT7"), ":"), list(list("ReporterIon"), " object for the TMT ", " 6-plex set plus the isobaric tag. Load with ", list("data(TMT6)"), ". ")), " ", " ")

Seealso

TMT6 or iTRAQ4 for readily available examples.

Author

Laurent Gatto lg390@cam.ac.uk

References

Ross PL, Huang YN, Marchese JN, Williamson B, Parker K, Hattan S, Khainovski N, Pillai S, Dey S, Daniels S, Purkayastha S, Juhasz P, Martin S, Bartlet-Jones M, He F, Jacobson A, Pappin DJ. "Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents." Mol Cell Proteomics , 2004 Dec;3(12):1154-69. Epub 2004 Sep 22. PubMed PMID: 15385600.

Thompson A, Sch" a fer J, Kuhn K, Kienle S, Schwarz J, Schmidt G, Neumann T, Johnstone R, Mohammed AK, Hamon C. "Tandem mass tags: a novel quantification strategy for comparative analysis of complex protein mixtures by MS/MS." Anal Chem. 2003 Apr 15;75(8):1895-904. Erratum in: Anal Chem. 2006 Jun 15;78(12):4235. Mohammed, A Karim A [added] and Anal Chem. 2003 Sep 15;75(18):4942. Johnstone, R [added]. PubMed PMID: 12713048.

Examples

## Code used for the iTRAQ4 set
ri <- new("ReporterIons",
description="4-plex iTRAQ",
name="iTRAQ4",
reporterNames=c("iTRAQ4.114","iTRAQ4.115",
"iTRAQ4.116","iTRAQ4.117"),
mz=c(114.1,115.1,116.1,117.1),
col=c("red","green","blue","yellow"),
width=0.05)
ri
reporterNames(ri)
ri[1:2]

List of Spectrum objects along with annotations

Description

Spectra objects allow to collect one or more Spectrum object(s) ( Spectrum1 or Spectrum2 ) in a list -like structure with the possibility to add arbitrary annotations to each individual Spectrum object. These can be accessed/set with the mcols() method.

Spectra objects can be created with the Spectra function.

Functions to access the individual spectra's attributes are available (listed below).

writeMgfData exports a Spectra object to a file in MGF format. All metadata columns present in mcols are exported as additional fields with the capitalized column names used as field names (see examples below).

Usage

Spectra(..., elementMetadata = NULL)
list(list("mz"), list("Spectra"))(object)
list(list("intensity"), list("Spectra"))(object)
list(list("rtime"), list("Spectra"))(object)
list(list("precursorMz"), list("Spectra"))(object)
list(list("precursorCharge"), list("Spectra"))(object)
list(list("precScanNum"), list("Spectra"))(object)
list(list("precursorIntensity"), list("Spectra"))(object)
list(list("acquisitionNum"), list("Spectra"))(object)
list(list("scanIndex"), list("Spectra"))(object)
list(list("peaksCount"), list("Spectra,ANY"))(object)
list(list("msLevel"), list("Spectra"))(object)
list(list("tic"), list("Spectra"))(object)
list(list("ionCount"), list("Spectra"))(object)
list(list("collisionEnergy"), list("Spectra"))(object)
list(list("fromFile"), list("Spectra"))(object)
list(list("polarity"), list("Spectra"))(object)
list(list("smoothed"), list("Spectra"))(object)
list(list("isEmpty"), list("Spectra"))(x)
list(list("centroided"), list("Spectra"))(object)
list(list("isCentroided"), list("Spectra"))(object)
list(list("writeMgfData"), list("Spectra"))(object, con = "spectra.mgf",
  COM = NULL, TITLE = NULL)
list(list("clean"), list("Spectra"))(object, all = FALSE, msLevel. = msLevel.,
  ...)
list(list("removePeaks"), list("Spectra"))(object, t, msLevel., ...)
list(list("filterMz"), list("Spectra"))(object, mz, msLevel., ...)
list(list("pickPeaks"), list("Spectra"))(object, halfWindowSize = 3L,
  method = c("MAD", "SuperSmoother"), SNR = 0L, refineMz = c("none",
  "kNeighbors", "kNeighbours", "descendPeak"), ...)
list(list("smooth"), list("Spectra"))(x, method = c("SavitzkyGolay",
  "MovingAverage"), halfWindowSize = 2L, ...)
list(list("filterMsLevel"), list("Spectra"))(object, msLevel.)

Arguments

Details

Spectra inherits all methods from the SimpleList class of the S4Vectors package. This includes lapply and other data manipulation and subsetting operations.

Author

Johannes Rainer

Examples

## Create from Spectrum objects
sp1 <- new("Spectrum1", mz = c(1, 2, 4), intensity = c(4, 5, 2))
sp2 <- new("Spectrum2", mz = c(1, 2, 3, 4), intensity = c(5, 3, 2, 5),
precursorMz = 2)

spl <- Spectra(sp1, sp2)
spl
spl[[1]]

## Add also metadata columns
mcols(spl)$id <- c("a", "b")
mcols(spl)

## Create a Spectra with metadata
spl <- Spectra(sp1, sp2, elementMetadata = DataFrame(id = c("a", "b")))

mcols(spl)
mcols(spl)$id

## Extract the mz values for the individual spectra
mz(spl)

## Extract the intensity values for the individual spectra
intensity(spl)

## Extract the retention time values for the individual spectra
rtime(spl)

## Extract the precursor m/z of each spectrum.
precursorMz(spl)

## Extract the precursor charge of each spectrum.
precursorCharge(spl)

## Extract the precursor scan number for each spectrum.
precScanNum(spl)

## Extract the precursor intensity of each spectrum.
precursorIntensity(spl)

## Extract the acquisition number of each spectrum.
acquisitionNum(spl)

## Extract the scan index of each spectrum.
scanIndex(spl)

## Get the number of peaks per spectrum.
peaksCount(spl)

## Get the MS level of each spectrum.
msLevel(spl)

## Get the total ion current for each spectrum.
tic(spl)

## Get the total ion current for each spectrum.
ionCount(spl)

## Extract the collision energy for each spectrum.
collisionEnergy(spl)

## Extract the file index for each spectrum.
fromFile(spl)

## Get the polarity for each spectrum.
polarity(spl)

## Whether spectra are smoothed (i.e. processed with the `smooth`
## function).
smoothed(spl)

## Are spectra empty (i.e. contain no peak data)?
isEmpty(spl)

## Do the spectra contain centroided data?
centroided(spl)

## Do the spectra contain centroided data? Whether spectra are centroided
## is estimated from the peak data.
isCentroided(spl)

## Export the spectrum list to a MGF file. Values in metadata columns are
## exported as additional field for each spectrum.
tmpf <- tempfile()
writeMgfData(spl, tmpf)

## Evaluate the written output. The ID of each spectrum (defined in the
## "id" metadata column) is exported as field "ID".
readLines(tmpf)

## Set mcols to NULL to avoid export of additional data fields.
mcols(spl) <- NULL
file.remove(tmpf)

writeMgfData(spl, tmpf)
readLines(tmpf)

## Filter the object by MS level
filterMsLevel(spl, msLevel. = 1)

The "Spectrum1" Class for MS1 Spectra

Description

Spectrum1 extends the " class and introduces an MS1 specific attribute in addition to the slots in " . Spectrum1 instances are not created directly but are contained in the assayData slot of an " .

Seealso

Virtual super-class " , " for MS2 spectra and " for a full experiment container.

Author

Laurent Gatto lg390@cam.ac.uk

The "Spectrum2" Class for MSn Spectra

Description

Spectrum2 extends the " class and introduces several MS2 specific attributes in addition to the slots in " . Since version 1.99.2, this class is used for any MS levels > 1. Spectrum2 are not created directly but are contained in the assayData slot of an " .

In version 1.19.12, the polarity slot had been added to the " class (previously in " ). Hence, "Spectrum2" objects created prior to this change will not be valid anymore, since they will miss the polarity slots. Object can be appropriately updated using the updateObject method.

Seealso

Virtual super-class " , " for MS1 spectra and " for a full experiment container.

Author

Laurent Gatto lg390@cam.ac.uk

The "Spectrum" Class

Description

Virtual container for spectrum data common to all different types of spectra. A Spectrum object can not be directly instanciated. Use " and " instead.

In version 1.19.12, the polarity slot has been added to this class (previously in " ).

Seealso

Instaciable sub-classes " and " for MS1 and MS2 spectra.

Note

This is a virtual class and can not be instanciated directly.

Author

Laurent Gatto lg390@cam.ac.uk

TMT 6/10-plex sets

Description

This instance of class " corresponds to the TMT 6-plex set, i.e the 126, 127, 128, 129, 130 and 131 isobaric tags. In the TMT7 data set, an unfragmented tag, i.e reporter and attached isobaric tag, is also included at MZ 229. A second TMT6b has slightly different values.

The TMT10 instance corresponds to the 10-plex version. There are spectific HCD ( TMT10HCD , same as TMT10 ) and ETD ( TMT10ETD ) sets.

These objects are used to plot the reporter ions of interest in an MSMS spectra (see " ) as well as for quantification (see quantify ).

Usage

TMT6
TMT6b
TMT7
TMT7b
TMT10
TMT10ETD
TMT10HCD
TMT11
TMT11HCD

Seealso

iTRAQ4 .

References

Thompson A, Sch" a fer J, Kuhn K, Kienle S, Schwarz J, Schmidt G, Neumann T, Johnstone R, Mohammed AK, Hamon C. "Tandem mass tags: a novel quantification strategy for comparative analysis of complex protein mixtures by MS/MS." Anal Chem. 2003 Apr 15;75(8):1895-904. Erratum in: Anal Chem. 2006 Jun 15;78(12):4235. Mohammed, A Karim A [added] and Anal Chem. 2003 Sep 15;75(18):4942. Johnstone, R [added]. PubMed PMID: 12713048.

Examples

TMT6
TMT6[1:2]

TMT10

newReporter <- new("ReporterIons",
description="an example",
name="my reporter ions",
reporterNames=c("myrep1","myrep2"),
mz=c(121,122),
col=c("red","blue"),
width=0.05)
newReporter

Adds Identification Data

Description

These methods add identification data to a raw MS experiment (an " object) or to quantitative data (an " object). The identification data needs to be available as a mzIdentML file (and passed as filenames, or directly as identification object) or, alternatively, can be passed as an arbitrary data.frame . See details in the Methods section.

Details

The featureData slots in a " or a " instance provides only one row per MS2 spectrum but the identification is not always bijective. Prior to addition, the identification data is filtered as documented in the filterIdentificationDataFrame function: (1) only PSMs matching the regular (non-decoy) database are retained; (2) PSMs of rank greater than 1 are discarded; and (3) only proteotypic peptides are kept.

If after filtering, more then one PSM per spectrum are still present, these are combined (reduced, see reduce,data.frame-method ) into a single row and separated by a semi-colon. This has as side-effect that feature variables that are being reduced are converted to characters. See the reduce manual page for examples.

See also the section about identification data in the MSnbase-demo vignette for details and additional examples.

After addition of the identification data, new feature variables are created. The column nprot contains the number of members in the protein group; the columns accession and description contain a semicolon separated list of all matches. The columns npsm.prot and npep.prot represent the number of PSMs and peptides that were matched to a particular protein group. The column npsm.pep indicates how many PSMs were attributed to a peptide (as defined by its sequence pepseq ). All these values are re-calculated after filtering and reduction.

Seealso

filterIdentificationDataFrame for the function that filters identification data, readMzIdData to read the identification data as a unfiltered data.frame and reduce,data.frame-method to reduce it to a data.frame that contains only unique PSMs per row.

Author

Sebastian Gibb mail@sebastiangibb.de and Laurent Gatto

Examples

## find path to a mzXML file
quantFile <- dir(system.file(package = "MSnbase", dir = "extdata"),
full.name = TRUE, pattern = "mzXML$")
## find path to a mzIdentML file
identFile <- dir(system.file(package = "MSnbase", dir = "extdata"),
full.name = TRUE, pattern = "dummyiTRAQ.mzid")

## create basic MSnExp
msexp <- readMSData(quantFile)

## add identification information
msexp <- addIdentificationData(msexp, identFile)

## access featureData
fData(msexp)

idSummary(msexp)

Identify aggregation outliers

Description

This function evaluates the variability within all protein group of an MSnSet . If a protein group is composed only of a single feature, NA is returned.

Usage

aggvar(object, groupBy, fun)

Arguments

Details

This function can be used to identify protein groups with incoherent feature (petides or PSMs) expression patterns. Using max as a function, one can identify protein groups with single extreme outliers, such as, for example, a mis-identified peptide that was erroneously assigned to that protein group. Using mean identifies more systematic inconsistencies where, for example, the subsets of peptide (or PSM) feautres correspond to proteins with different expression patterns.

Value

A matrix providing the number of features per protein group ( nb_feats column) and the aggregation summarising distance ( agg_dist column).

Seealso

combineFeatures to combine PSMs quantitation into peptides and/or into proteins.

Author

Laurent Gatto

Examples

library("pRolocdata")
data(hyperLOPIT2015ms3r1psm)
groupBy <- "Protein.Group.Accessions"
res1 <- aggvar(hyperLOPIT2015ms3r1psm, groupBy, fun = max)
res2 <- aggvar(hyperLOPIT2015ms3r1psm, groupBy, fun = mean)
par(mfrow = c(1, 3))
plot(res1, log = "y", main = "Single outliers (max)")
plot(res2, log = "y", main = "Overall inconsistency (mean)")
plot(res1[, "agg_dist"], res2[, "agg_dist"],
xlab = "max", ylab = "mean")

Generate an average MSnSet

Description

Given a list of MSnSet instances, typically representing replicated experiments, the function returns an average MSnSet .

Usage

averageMSnSet(x, avg = function(x) mean(x, na.rm = TRUE), disp = npcv)

Arguments

Details

This function is aimed at facilitating the visualisation of replicated experiments and should not be used as a replacement for a statistical analysis.

The samples of the instances to be averaged must be identical but can be in a different order (they will be reordered by default). The features names of the result will correspond to the union of the feature names of the input MSnSet instances. Each average value will be computed by the avg function and the dispersion of the replicated measurements will be estimated by the disp function. These dispersions will be stored as a data.frame in the feature metadata that can be accessed with fData(.)$disp . Similarly, the number of missing values that were present when average (and dispersion) were computed are available in fData(.)$disp .

Currently, the feature metadata of the returned object corresponds the the feature metadata of the first object in the list (augmented with the missing value and dispersion values); the metadata of the features that were missing in this first input are missing (i.e. populated with NA s). This may change in the future.

Value

A new average MSnSet .

Seealso

compfnames to compare MSnSet feature names.

Author

Laurent Gatto

Examples

library("pRolocdata")
## 3 replicates from Tan et al. 2009
data(tan2009r1)
data(tan2009r2)
data(tan2009r3)
x <- MSnSetList(list(tan2009r1, tan2009r2, tan2009r3))
avg <- averageMSnSet(x)
dim(avg)
head(exprs(avg))
head(fData(avg)$nNA)
head(fData(avg)$disp)
## using the standard deviation as measure of dispersion
avg2 <-averageMSnSet(x, disp = sd)
head(fData(avg2)$disp)
## keep only complete observations, i.e proteins
## that had 0 missing values for all samples
sel <- apply(fData(avg)$nNA, 1 , function(x) all(x == 0))
avg <- avg[sel, ]
disp <- rowMax(fData(avg)$disp)
library("pRoloc")
setStockcol(paste0(getStockcol(), "AA"))
plot2D(avg, cex = 7.7 * disp)
title(main = paste("Dispersion: non-parametric CV",
paste(round(range(disp), 3), collapse = " - ")))

Bin 'MSnExp' or 'Spectrum' instances

Description

This method aggregates individual spectra ( Spectrum instances) or whole experiments ( MSnExp instances) into discrete bins. All intensity values which belong to the same bin are summed together.

Seealso

clean , pickPeaks , smooth , removePeaks and trimMz for other spectra processing methods.

Author

Sebastian Gibb mail@sebastiangibb.de

Examples

s <- new("Spectrum2", mz=1:10, intensity=1:10)
intensity(s)
intensity(bin(s, binSize=2))

data(itraqdata)
sum(peaksCount(itraqdata))
itraqdata2 <- bin(itraqdata, binSize=2)
sum(peaksCount(itraqdata2))
processingData(itraqdata2)

Calculate ions produced by fragmentation.

Description

These method calculates a-, b-, c-, x-, y- and z-ions produced by fragmentation.

Arguments

Author

Sebastian Gibb mail@sebastiangibb.de

Examples

## find path to a mzXML file
file <- dir(system.file(package = "MSnbase", dir = "extdata"),
full.name = TRUE, pattern = "mzXML$")

## create basic MSnExp
msexp <- readMSData(file, centroided = FALSE)

## centroid them
msexp <- pickPeaks(msexp)

## calculate fragments for ACE with default modification
calculateFragments("ACE", modifications=c(C=57.02146))

## calculate fragments for ACE with an addition N-terminal modification
calculateFragments("ACE", modifications=c(C=57.02146, Nterm=229.1629))

## calculate fragments for ACE without any modifications
calculateFragments("ACE", modifications=NULL)

calculateFragments("VESITARHGEVLQLRPK",
type=c("a", "b", "c", "x", "y", "z"),
z=1:2)

calculateFragments("VESITARHGEVLQLRPK", msexp[[1]])

## neutral loss
defaultNeutralLoss()

## disable water loss on the C terminal
defaultNeutralLoss(disableWaterLoss="Cterm")

## real example
calculateFragments("PQR")
calculateFragments("PQR",
neutralLoss=defaultNeutralLoss(disableWaterLoss="Cterm"))
calculateFragments("PQR",
neutralLoss=defaultNeutralLoss(disableAmmoniaLoss="Q"))

## disable neutral loss completely
calculateFragments("PQR", neutralLoss=NULL)

Extract chromatogram object(s)

Description

The chromatogram method extracts chromatogram(s) from an MSnExp or OnDiskMSnExp object. Depending on the provided parameters this can be a total ion chromatogram (TIC), a base peak chromatogram (BPC) or an extracted ion chromatogram (XIC) extracted from each sample/file.

Usage

list(list("chromatogram"), list("MSnExp"))(object, rt, mz, aggregationFun = "sum",
  missing = NA_real_, msLevel = 1L, BPPARAM = bpparam())

Arguments

Details

Arguments rt and mz allow to specify the MS data slice from which the chromatogram should be extracted. The parameter aggregationSum allows to specify the function to be used to aggregate the intensities across the mz range for the same retention time. Setting aggregationFun = "sum" would e.g. allow to calculate the total ion chromatogram (TIC), aggregationFun = "max" the base peak chromatogram (BPC). The length of the extracted Chromatogram object, i.e. the number of available data points, corresponds to the number of scans/spectra measured in the specified retention time range. If in a specific scan (for a give retention time) no signal was measured in the specified mz range, a NA_real_ is reported as intensity for the retention time (see Notes for more information). This can be changed using the missing parameter.

By default or if mz and/or rt are numeric vectors, the function extracts one Chromatogram object for each file in the MSnExp or OnDiskMSnExp object. Providing a numeric matrix with argument mz or rt enables to extract multiple chromatograms per file, one for each row in the matrix. If the number of columns of mz or rt are not equal to 2, range is called on each row of the matrix.

Value

chromatogram returns a Chromatograms object with the number of columns corresponding to the number of files in object and number of rows the number of specified ranges (i.e. number of rows of matrices provided with arguments mz and/or rt ). The featureData of the returned object contains columns "mzmin" and "mzmax" with the values from input argument mz (if used) and "rtmin" and "rtmax" if the input argument rt was used.

Seealso

Chromatogram and Chromatograms for the classes that represent single and multiple chromatograms.

Author

Johannes Rainer

Examples

## Read a test data file.
library(msdata)
f <- c(system.file("microtofq/MM14.mzML", package = "msdata"),
system.file("microtofq/MM8.mzML", package = "msdata"))

## Read the data as an MSnExp
msd <- readMSData(f, msLevel = 1)

## Extract the total ion chromatogram for each file:
tic <- chromatogram(msd)

tic

## Extract the TIC for the second file:
tic[1, 2]

## Plot the TIC for the first file
plot(rtime(tic[1, 1]), intensity(tic[1, 1]), type = "l",
xlab = "rtime", ylab = "intensity", main = "TIC")

## Extract chromatograms for a MS data slices defined by retention time
## and mz ranges.
rtr <- rbind(c(10, 60), c(280, 300))
mzr <- rbind(c(140, 160), c(300, 320))
chrs <- chromatogram(msd, rt = rtr, mz = mzr)

## Each row of the returned Chromatograms object corresponds to one mz-rt
## range. The Chromatogram for the first range in the first file is empty,
## because the retention time range is outside of the file's rt range:
chrs[1, 1]

## The mz and/or rt ranges used are provided as featureData of the object
fData(chrs)

## The mz method can be used to extract the m/z ranges directly
mz(chrs)

## Also the Chromatogram for the second range in the second file is empty
chrs[2, 2]

## Get the extracted chromatogram for the first range in the second file
chr <- chrs[1, 2]
chr

plot(rtime(chr), intensity(chr), xlab = "rtime", ylab = "intensity")

Clean 'MSnExp', 'Spectrum' or 'Chromatogram' instances

Description

This method cleans out individual spectra ( Spectrum instances), chromatograms ( Chromatogram instances) or whole experiments ( MSnExp instances) of 0-intensity peaks. Unless all is set to FALSE , original 0-intensity values are retained only around peaks. If more than two 0's were separating two peaks, only the first and last ones, those directly adjacent to the peak ranges are kept. If two peaks are separated by only one 0-intensity value, it is retained. An illustrative example is shown below.

Seealso

removePeaks and trimMz for other spectra processing methods.

Author

Laurent Gatto lg390@cam.ac.uk

Examples

int <- c(1,0,0,0,0,0,0,0,1,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,1,0,0,0)
sp1 <- new("Spectrum2",
intensity=int,
mz=1:length(int))
sp2 <- clean(sp1) ## default is all=FALSE
intensity(sp1)
intensity(sp2)
intensity(clean(sp1, all = TRUE))

mz(sp1)
mz(sp2)
mz(clean(sp1, all = TRUE))

data(itraqdata)
itraqdata2 <- clean(itraqdata)
sum(peaksCount(itraqdata))
sum(peaksCount(itraqdata2))
processingData(itraqdata2)

## Create a simple Chromatogram object
chr <- Chromatogram(rtime = 1:12,
intensity = c(0, 0, 20, 0, 0, 0, 123, 124343, 3432, 0, 0, 0))

## Remove 0-intensity values keeping those adjacent to peaks
chr <- clean(chr)
intensity(chr)

## Remove all 0-intensity values
chr <- clean(chr, all = TRUE)
intensity(chr)

## Clean a Chromatogram with NAs.
chr <- Chromatogram(rtime = 1:12,
intensity = c(0, 0, 20, NA, NA, 0, 123, 124343, 3432, 0, 0, 0))
chr <- clean(chr, all = FALSE, na.rm = TRUE)
intensity(chr)

Combines features in an MSnSet object

Description

This function combines the features in an " instance applying a summarisation function (see fun argument) to sets of features as defined by a factor (see fcol argument). Note that the feature names are automatically updated based on the groupBy parameter.

The coefficient of variations are automatically computed and collated to the featureData slot. See cv and cv.norm arguments for details.

If NA values are present, a message will be shown. Details on how missing value impact on the data aggregation are provided below.

Usage

combineFeatures(object, groupBy, method = c("mean", "median",
"weighted.mean", "sum", "medpolish", "robust", "iPQF", "NTR"), fcol,
redundancy.handler = c("unique", "multiple"), cv = TRUE, cv.norm =
"sum", verbose = isMSnbaseVerbose(), fun, ...)

Arguments

Details

Missing values have different effect based on the aggregation method employed, as detailed below. See also examples below.

• When using either "sum" , "mean" , "weighted.mean" or "median" , any missing value will be propagated at the higher level. If na.rm = TRUE is used, then the missing value will be ignored.

• Missing values will result in an error when using "medpolish" , unless na.rm = TRUE is used.

• When using robust summarisation ( "robust" ), individual missing values are excluded prior to fitting the linear model by robust regression. To remove all values in the feature containing the missing values, use filterNA .

• The "iPQF" method will fail with an error if missing value are present, which will have to be handled explicitly. See below.

More generally, missing values often need dedicated handling such as filtering (see filterNA ) or imputation (see impute ).

Value

A new " instance is returned with ncol (i.e. number of samples) is unchanged, but nrow (i.e. the number od features) is now equals to the number of levels in groupBy . The feature metadata ( featureData slot) is updated accordingly and only the first occurrence of a feature in the original feature meta-data is kept.

Seealso

featureCV to calculate coefficient of variation, nFeatures to document the number of features per group in the feature data, and the aggvar to explore variability within protein groups.

iPQF for iPQF summarisation.

NTR for normalisation to reference summarisation.

Author

Laurent Gatto lg390@cam.ac.uk with contributions from Martina Fischer for iPQF and Ludger Goeminne, Adriaan Sticker and Lieven Clement for robust.

References

iPQF: a new peptide-to-protein summarization method using peptide spectra characteristics to improve protein quantification. Fischer M, Renard BY. Bioinformatics. 2016 Apr 1;32(7):1040-7. doi:10.1093/bioinformatics/btv675. Epub 2015 Nov

20. PubMed PMID:26589272.

Examples

data(msnset)
msnset <- msnset[11:15, ]
exprs(msnset)

## arbitrary grouping into two groups
grp <- as.factor(c(1, 1, 2, 2, 2))
msnset.comb <- combineFeatures(msnset, grp, "sum")
dim(msnset.comb)
exprs(msnset.comb)
fvarLabels(msnset.comb)

## grouping with a list
grpl <- list(c("A", "B"), "A", "A", "C", c("C", "B"))
## optional naming
names(grpl) <- featureNames(msnset)
exprs(combineFeatures(msnset, grpl, method = "sum", redundancy.handler = "unique"))
exprs(combineFeatures(msnset, grpl, method = "sum", redundancy.handler = "multiple"))

## missing data
exprs(msnset)[4, 4] <-
exprs(msnset)[2, 2] <- NA
exprs(msnset)
## NAs propagate in the 115 and 117 channels
exprs(combineFeatures(msnset, grp, "sum"))
## NAs are removed before summing
exprs(combineFeatures(msnset, grp, "sum", na.rm = TRUE))

## using iPQF
data(msnset2)
anyNA(msnset2)
res <- combineFeatures(msnset2,
groupBy = fData(msnset2)$accession,
redundancy.handler = "unique",
method = "iPQF",
low.support.filter = FALSE,
ratio.calc = "sum",
method.combine = FALSE)
head(exprs(res))

## using robust summarisation
data(msnset) ## reset data
msnset <- log(msnset, 2) ## log2 transform

## Feature X46, in the ENO protein has one missig value
which(is.na(msnset), arr.ind = dim(msnset))
exprs(msnset["X46", ])
## Only the missing value in X46 and iTRAQ4.116 will be ignored
res <- combineFeatures(msnset,
fcol = "ProteinAccession",
method = "robust")
tail(exprs(res))

msnset2 <- filterNA(msnset) ## remove features with missing value(s)
res2 <- combineFeatures(msnset2,
fcol = "ProteinAccession",
method = "robust")
## Here, the values for ENO are different because the whole feature
## X46 that contained the missing value was removed prior to fitting.
tail(exprs(res2))

Combine Spectra

Description

combineSpectra combines spectra in a MSnExp or Spectra object applying the summarization function fun to sets of spectra defined by a factor ( fcol parameter). The resulting combined spectrum for each set contains metadata information (present in mcols and all spectrum information other than mz and intensity ) from the first spectrum in each set.

Usage

list(list("combineSpectra"), list("Spectra"))(object, fcol, method = meanMzInts,
  fun, ...)

Arguments

Value

A Spectra or MSnExp object with combined spectra. Metadata ( mcols ) and all spectrum attributes other than mz and intensity are taken from the first Spectrum in each set.

Seealso

meanMzInts() for a function to combine spectra.

Author

Johannes Rainer, Laurent Gatto

Examples

set.seed(123)
mzs <- seq(1, 20, 0.1)
ints1 <- abs(rnorm(length(mzs), 10))
ints1[11:20] <- c(15, 30, 90, 200, 500, 300, 100, 70, 40, 20) # add peak
ints2 <- abs(rnorm(length(mzs), 10))
ints2[11:20] <- c(15, 30, 60, 120, 300, 200, 90, 60, 30, 23)
ints3 <- abs(rnorm(length(mzs), 10))
ints3[11:20] <- c(13, 20, 50, 100, 200, 100, 80, 40, 30, 20)

## Create the spectra.
sp1 <- new("Spectrum1", mz = mzs + rnorm(length(mzs), sd = 0.01),
intensity = ints1, rt = 1)
sp2 <- new("Spectrum1", mz = mzs + rnorm(length(mzs), sd = 0.01),
intensity = ints2, rt = 2)
sp3 <- new("Spectrum1", mz = mzs + rnorm(length(mzs), sd = 0.009),
intensity = ints3, rt = 3)

spctra <- Spectra(sp1, sp2, sp3,
elementMetadata = DataFrame(idx = 1:3, group = c("b", "a", "a")))

## Combine the spectra reporting the maximym signal
res <- combineSpectra(spctra, mzd = 0.05, intensityFun = max)
res

## All values other than m/z and intensity are kept from the first spectrum
rtime(res)

## Plot the individual and the merged spectrum
par(mfrow = c(2, 1), mar = c(4.3, 4, 1, 1))
plot(mz(sp1), intensity(sp1), xlim = range(mzs[5:25]), type = "h", col = "red")
points(mz(sp2), intensity(sp2), type = "h", col = "green")
points(mz(sp3), intensity(sp3), type = "h", col = "blue")
plot(mz(res[[1]]), intensity(res[[1]]), type = "h",
col = "black", xlim = range(mzs[5:25]))

## Combine spectra in two sets.
res <- combineSpectra(spctra, fcol = "group", mzd = 0.05)
res

rtime(res)

## Plot the individual and the merged spectra
par(mfrow = c(3, 1), mar = c(4.3, 4, 1, 1))
plot(mz(sp1), intensity(sp1), xlim = range(mzs[5:25]), type = "h", col = "red")
points(mz(sp2), intensity(sp2), type = "h", col = "green")
points(mz(sp3), intensity(sp3), type = "h", col = "blue")
plot(mz(res[[1]]), intensity(res[[1]]), xlim = range(mzs[5:25]), type = "h",
col = "black")
plot(mz(res[[2]]), intensity(res[[2]]), xlim = range(mzs[5:25]), type = "h",
col = "black")

Combine signal from consecutive spectra of LCMS experiments

Description

combineSpectraMovingWindow combines signal from consecutive spectra within a file. The resulting MSnExp has the same total number of spectra than the original object, but with each individual's spectrum information representing aggregated data from the original spectrum and its neighboring spectra. This is thus equivalent with a smoothing of the data in retention time dimension.

Note that the function returns always a MSnExp object, even if x was an OnDiskMSnExp object.

Usage

combineSpectraMovingWindow(x, halfWindowSize = 1L,
  intensityFun = base::mean, mzd = NULL, timeDomain = FALSE,
  weighted = FALSE, BPPARAM = bpparam())

Arguments

Details

The method assumes same ions being measured in consecutive scans (i.e. LCMS data) and thus combines their signal which can increase the increase the signal to noise ratio.

Intensities (and m/z values) for signals with the same m/z value in consecutive scans are aggregated using the intensityFun . m/z values of intensities from consecutive scans will never be exactly identical, even if they represent signal from the same ion. The function determines thus internally a similarity threshold based on differences between m/z values within and between spectra below which m/z values are considered to derive from the same ion. For robustness reasons, this threshold is estimated on the 100 spectra with the largest number of m/z - intensity pairs (i.e. mass peaks).

See meanMzInts() for details.

Parameter timeDomain : by default, m/z-intensity pairs from consecutive scans to be aggregated are defined based on the square root of the m/z values. This is because it is highly likely that in all QTOF MS instruments data is collected based on a timing circuit (with a certain variance) and m/z values are later derived based on the relationship t = k * sqrt(m/z) . Differences between individual m/z values will thus be dependent on the actual m/z value causing both the difference between m/z values and their scattering being different in the lower and upper m/z range. Determining m/z values to be combined on the sqrt(mz) reduces this dependency. For non-QTOF MS data timeDomain = FALSE might be used instead.

Value

MSnExp with the same number of spectra than x .

Seealso

meanMzInts() for the function combining spectra provided in a list .

estimateMzScattering() for a function to estimate m/z value scattering in consecutive spectra.

Note

The function has to read all data into memory for the spectra combining and thus the memory requirements of this function are high, possibly preventing its usage on large experimental data. In these cases it is suggested to perform the combination on a per-file basis and save the results using the writeMSData() function afterwards.

Author

Johannes Rainer, Sigurdur Smarason

Examples

library(MSnbase)
library(msdata)

## Read a profile-mode LC-MS data file.
fl <- dir(system.file("sciex", package = "msdata"), full.names = TRUE)[1]
od <- readMSData(fl, mode = "onDisk")

## Subset the object to the retention time range that includes the signal
## for proline. This is done for performance reasons.
rtr <- c(165, 175)
od <- filterRt(od, rtr)

## Combine signal from neighboring spectra.
od_comb <- combineSpectraMovingWindow(od)

## The combined spectra have the same number of spectra, same number of
## mass peaks per spectra, but the signal is larger in the combined object.
length(od)
length(od_comb)

peaksCount(od)
peaksCount(od_comb)

## Comparing the chromatographic signal for proline (m/z ~ 116.0706)
## before and after spectra data combination.
mzr <- c(116.065, 116.075)
chr <- chromatogram(od, rt = rtr, mz = mzr)
chr_comb <- chromatogram(od_comb, rt = rtr, mz = mzr)

par(mfrow = c(1, 2))
plot(chr)
plot(chr_comb)
## Chromatographic data is "smoother" after combining.

Keep only common feature names

Description

Subsets MSnSet instances to their common feature names.

Usage

commonFeatureNames(x, y)

Arguments

Value

An linkS4class{MSnSetList} composed of the input MSnSet containing only common features in the same order. The names of the output are either the names of the x and y input variables or the names of x if a list is provided.

Author

Laurent Gatto

Examples

library("pRolocdata")
data(tan2009r1)
data(tan2009r2)
cmn <- commonFeatureNames(tan2009r1, tan2009r2)
names(cmn)
## as a named list
names(commonFeatureNames(list(a = tan2009r1, b = tan2009r2)))
## without message
suppressMessages(cmn <- commonFeatureNames(tan2009r1, tan2009r2))
## more than 2 instance
data(tan2009r3)
cmn <- commonFeatureNames(list(tan2009r1, tan2009r2, tan2009r3))
length(cmn)

Compare two MSnSets

Description

Compares two MSnSet instances. The qual and processingData slots are generally omitted.

Usage

compareMSnSets(x, y, qual = FALSE, proc = FALSE)

Arguments

Value

A logical

Author

Laurent Gatto

Compare Spectra of an 'MSnExp' or 'Spectrum' instances

Description

This method compares spectra ( Spectrum instances) pairwise or all spectra of an experiment ( MSnExp instances). Currently the comparison is based on the number of common peaks fun = "common" , the Pearson correlation fun = "cor" , the dot product fun = "dotproduct" or a user-defined function.

For fun = "common" the tolerance (default 25e-6 ) can be set and the tolerance can be defined to be relative (default relative = TRUE ) or absolute ( relative = FALSE ). To compare spectra with fun = "cor" and fun = "dotproduct" , the spectra need to be binned. The binSize argument (in Dalton) controls the binning precision. Please see bin for details.

Instead of these three predefined functions for fun a user-defined comparison function can be supplied. This function takes two list("Spectrum") objects as the first two arguments and list() as third argument. The function must return a single numeric value. See the example section.

Seealso

bin , clean , pickPeaks , smooth , removePeaks and trimMz for other spectra processing methods.

Author

Sebastian Gibb mail@sebastiangibb.de

References

Stein, S. E., & Scott, D. R. (1994). Optimization and testing of mass spectral library search algorithms for compound identification. Journal of the American Society for Mass Spectrometry, 5(9), 859-866. doi: https://doi.org/10.1016/1044-0305(94)87009-8

Lam, H., Deutsch, E. W., Eddes, J. S., Eng, J. K., King, N., Stein, S. E. and Aebersold, R. (2007) Development and validation of a spectral library searching method for peptide identification from MS/MS. Proteomics, 7: 655-667. doi: https://doi.org/10.1002/pmic.200600625

Examples

s1 <- new("Spectrum2", mz=1:10, intensity=1:10)
s2 <- new("Spectrum2", mz=1:10, intensity=10:1)
compareSpectra(s1, s2)
compareSpectra(s1, s2, fun="cor", binSize=2)
compareSpectra(s1, s2, fun="dotproduct")

## define our own (useless) comparison function (it is just a basic example)
equalLength <- function(x, y, ...) {
return(peaksCount(x)/(peaksCount(y)+.Machine$double.eps))
}
compareSpectra(s1, s2, fun=equalLength)
compareSpectra(s1, new("Spectrum2", mz=1:5, intensity=1:5), fun=equalLength)
compareSpectra(s1, new("Spectrum2"), fun=equalLength)

data(itraqdata)
compareSpectra(itraqdata[1:5], fun="cor")

Combine spectra to a consensus spectrum

Description

consensusSpectrum takes a list of spectra and combines them to a consensus spectrum containing mass peaks that are present in a user definable proportion of spectra.

Usage

consensusSpectrum(x, mzd, minProp = 0.5, intensityFun = base::max,
  ppm = 0, ...)

Arguments

Details

Peaks from spectra with a difference of their m/z being smaller than mzd are grouped into the same final mass peak with their intensities being aggregated with intensityFun . The m/z of the final mass peaks is calculated using a intensity-weighted mean of the m/z values from the individual mass peaks. Alternatively (or in addition) it is possible to perform an m/z dependent grouping of mass peaks with parameter ppm : mass peaks from different spectra with a difference in their m/z smaller than ppm of their m/z are grouped into the same final peak.

Seealso

Other spectra combination functions: meanMzInts

Author

Johannes Rainer

Examples

library(MSnbase)
## Create 3 example spectra.
sp1 <- new("Spectrum2", rt = 1, precursorMz = 1.41,
mz = c(1.2, 1.5, 1.8, 3.6, 4.9, 5.0, 7.8, 8.4),
intensity = c(10, 3, 140, 14, 299, 12, 49, 20))
sp2 <- new("Spectrum2", rt = 1.1, precursorMz = 1.4102,
mz = c(1.4, 1.81, 2.4, 4.91, 6.0, 7.2, 9),
intensity = c(3, 184, 8, 156, 12, 23, 10))
sp3 <- new("Spectrum2", rt = 1.2, precursorMz = 1.409,
mz = c(1, 1.82, 2.2, 3, 7.0, 8),
intensity = c(8, 210, 7, 101, 17, 8))
spl <- Spectra(sp1, sp2, sp3)

## Plot the spectra, each in a different color
par(mfrow = c(2, 1), mar = c(4.3, 4, 1, 1))
plot(mz(sp1), intensity(sp1), type = "h", col = "#ff000080", lwd = 2,
xlab = "m/z", ylab = "intensity", xlim = range(mz(spl)),
ylim = range(intensity(spl)))
points(mz(sp2), intensity(sp2), type = "h", col = "#00ff0080", lwd = 2)
points(mz(sp3), intensity(sp3), type = "h", col = "#0000ff80", lwd = 2)

cons <- consensusSpectrum(spl, mzd = 0.02, minProp = 2/3)

## Peaks of the consensus spectrum
mz(cons)
intensity(cons)

## Other Spectrum data is taken from the first Spectrum in the list
rtime(cons)
precursorMz(cons)

plot(mz(cons), intensity(cons), type = "h", xlab = "m/z", ylab = "intensity",
xlim = range(mz(spl)), ylim = range(intensity(spl)), lwd = 2)

MSnbase Deprecated and Defunct

Description

The function, class, or data object you have asked for has been deprecated or made defunct.

Deprecated:

Defunct: readMzXMLData , extractSpectra , writeMzTabData , makeMTD , makePEP , makePRT , NAnnotatedDataFrame class.

Estimate the m/z resolution of a spectrum

Description

estimateMzResolution estimates the m/z resolution of a profile-mode Spectrum (or of all spectra in an MSnExp or OnDiskMSnExp object. The m/z resolution is defined as the most frequent difference between a spectrum's m/z values.

Usage

list(list("estimateMzResolution"), list("MSnExp"))(object, ...)
list(list("estimateMzResolution"), list("Spectrum"))(object, ...)

Arguments

Value

numeric(1) with the m/z resolution. If called on a MSnExp or OnDiskMSnExp a list of m/z resolutions are returned (one for each spectrum).

Note

This assumes the data to be in profile mode and does not return meaningful results for centroided data.

The estimated m/z resolution depends on the number of ions detected in a spectrum, as some instrument don't measure (or report) signal if below a certain threshold.

Author

Johannes Rainer

Examples

## Load a profile mode example file
library(MSnbase)
library(msdata)
f <- proteomics(full.names = TRUE,
pattern = "TMT_Erwinia_1uLSike_Top10HCD_isol2_45stepped_60min_01.mzML.gz")

od <- readMSData(f, mode = "onDisk")

## Estimate the m/z resolution on the 3rd spectrum.
estimateMzResolution(od[[3]])

## Estimate the m/z resolution for each spectrum
mzr <- estimateMzResolution(od)

## plot the distribution of estimated m/z resolutions. The bimodal
## distribution represents the m/z resolution of the MS1 (first peak) and
## MS2 spectra (second peak).
plot(density(unlist(mzr)))

Estimate m/z scattering in consecutive scans

Description

Estimate scattering of m/z values (due to technical, instrument specific noise) for the same ion in consecutive scans of a LCMS experiment.

Usage

estimateMzScattering(x, halfWindowSize = 1L, timeDomain = FALSE)

Arguments

Details

The m/z values of the same ions in consecutive scans (spectra) of a LCMS run will not be identical. This random noise is expected to be smaller than the resolution of the MS instrument. The distribution of differences of m/z values from neighboring spectra is thus expected to be (at least) bi-modal with the first peak representing the above described random variation and the second (or largest) peak the m/z resolution. The m/z value of the first local minimum between these first two peaks in the distribution is returned as the m/z scattering .

Seealso

estimateMzResolution() for the function to estimate a profile-mode spectrum's m/z resolution from it's data.

Note

For timeDomain = TRUE the function does not return the estimated scattering of m/z values, but the scattering of sqrt(mz) values.

Author

Johannes Rainer

Examples

library(MSnbase)
library(msdata)
## Load a profile-mode LC-MS data file
f <- dir(system.file("sciex", package = "msdata"), full.names = TRUE)[1]
im <- readMSData(f, mode = "inMem", msLevel = 1L)

res <- estimateMzScattering(im)

## Plot the distribution of estimated m/z scattering
plot(density(unlist(res)))

## Compare the m/z resolution and m/z scattering of the spectrum with the
## most peaks
idx <- which.max(unlist(spectrapply(im, peaksCount)))

res[[idx]]
abline(v = res[[idx]], lty = 2)
estimateMzResolution(im[[idx]])
## As expected, the m/z scattering is much lower than the m/z resolution.

Noise Estimation for 'Spectrum' instances

Description

This method performs a noise estimation on individual spectra ( Spectrum instances). There are currently two different noise estimators, the Median Absolute Deviation ( method = "MAD" ) and Friedman's Super Smoother ( method = "SuperSmoother" ), as implemented in the MALDIquant::detectPeaks and MALDIquant::estimateNoise functions respectively.

Seealso

pickPeaks , and the underlying method in MALDIquant : estimateNoise .

Author

Sebastian Gibb mail@sebastiangibb.de

References

S. Gibb and K. Strimmer. 2012. MALDIquant: a versatile R package for the analysis of mass spectrometry data. Bioinformatics 28: 2270-2271. http://strimmerlab.org/software/maldiquant/

Examples

sp1 <- new("Spectrum1",
intensity = c(1:6, 5:1),
mz = 1:11,
centroided = FALSE)
estimateNoise(sp1, method = "SuperSmoother")

Calculate all ratio pairs

Description

Calculations all possible ratios for the assayData columns in an " . The function getRatios(x, log = FALSE) takes a matrix x as input and is used by exprsToRatios .

Examples

data(msnset)
pData(msnset)
head(exprs(msnset))
r <- exprsToRatios(msnset)
head(exprs(r))
pData(r)

Extracts precursor-specific spectra from an 'MSnExp' object

Description

Extracts the MSMS spectra that originate from the precursor(s) having the same MZ value as defined in the prec argument.

A warning will be issued of one or several of the precursor MZ values in prec are absent in the experiment precursor MZ values (i.e in precursorMz(object) ).

Author

Laurent Gatto lg390@cam.ac.uk

Examples

file <- dir(system.file(package="MSnbase",dir="extdata"),
full.name=TRUE,pattern="mzXML$")
aa <- readMSData(file,verbose=FALSE)
my.prec <- precursorMz(aa)[1]
my.prec
bb <- extractPrecSpectra(aa,my.prec)
precursorMz(bb)
processingData(bb)

Expand or merge feature variables

Description

The expandFeatureVars and mergeFeatureVars respectively expand and merge groups of feature variables. Using these functions, a set of columns in a feature data can be merged into a single new data.frame-column variables and a data.frame-column can be expanded into single feature columns. The original feature variables are removed.

Usage

expandFeatureVars(x, fcol, prefix)
mergeFeatureVars(x, fcol, fcol2)

Arguments

Value

An MSnSet for expanded (merged) feature variables.

Author

Laurent Gatto

Examples

library("pRolocdata")
data(hyperLOPIT2015)
fvarLabels(hyperLOPIT2015)
## Let's merge all svm prediction feature variables
(k <- grep("^svm", fvarLabels(hyperLOPIT2015), value = TRUE))
hl <- mergeFeatureVars(hyperLOPIT2015, fcol = k, fcol2 = "SVM")
fvarLabels(hl)
head(fData(hl)$SVM)

## Let's expand the new SVM into individual columns
hl2 <- expandFeatureVars(hl, "SVM")
fvarLabels(hl2)
## We can set the prefix manually
hl2 <- expandFeatureVars(hl, "SVM", prefix = "Expanded")
fvarLabels(hl2)
## If we don't want any prefix
hl2 <- expandFeatureVars(hl, "SVM", prefix = NULL)
fvarLabels(hl2)

Converts factors to strings

Description

This function produces the opposite as the stringsAsFactors argument in the data.frame or read.table functions; it converts factors columns to characters .

Usage

factorsAsStrings(x)

Arguments

Value

A data.frame where factors are converted to characters .

Author

Laurent Gatto

Examples

data(iris)
str(iris)
str(factorsAsStrings(iris))

Calculates coeffivient of variation for features

Description

This function calculates the column-wise coefficient of variation (CV), i.e. the ration between the standard deviation and the mean, for the features in an MSnSet . The CVs are calculated for the groups of features defined by groupBy . For groups defined by single features, NA is returned.

Usage

featureCV(x, groupBy, na.rm = TRUE, norm = "none", suffix = NULL)

Arguments

Value

A matrix of dimensions length(levels(groupBy)) by ncol(x) with the respecive CVs. The column names are formed by pasting CV. and the sample names of object x , possibly suffixed by .suffix .

Seealso

combineFeatures()

Author

Laurent Gatto and Sebastian Gibb

Examples

data(msnset)
msnset <- msnset[1:4]
gb <- factor(rep(1:2, each = 2))
featureCV(msnset, gb)
featureCV(msnset, gb, suffix = "2")

Fills up a vector

Description

This function replaces all the empty characters "" and/or NA s with the value of the closest preceding the preceding non- NA / "" element. The function is used to populate dataframe or matrice columns where only the cells of the first row in a set of partially identical rows are explicitly populated and the following are empty.

Usage

fillUp(x)

Arguments

Value

A vector as x with all empty characters "" and NA values replaced by the preceding non- NA / "" value.

Author

Laurent Gatto lg390@cam.ac.uk

Examples

d <- data.frame(protein=c("Prot1","","","Prot2","",""),
peptide=c("pep11","","pep12","pep21","pep22",""),
score=c(1:2,NA,1:3))
d
e <- apply(d,2,fillUp)
e
data.frame(e)
fillUp(d[,1])

Filter out unreliable PSMs.

Description

A function to filter out PSMs matching to the decoy database, of rank greater than one and matching non-proteotypic peptides.

Usage

filterIdentificationDataFrame(x, decoy = "isDecoy", rank = "rank",
  accession = "DatabaseAccess", spectrumID = "spectrumID",
  verbose = isMSnbaseVerbose())

Arguments

Details

The PSMs should be stored in a data.frame such as those produced by readMzIdData() . Note that this function should be called before calling the reduce method on a PSM data.frame .

Value

A new data.frame with filtered out peptides and with the same columns as the input x .

Author

Laurent Gatto

Format Retention Time

Description

Converts seconds to/from 'min:sec' format

Usage

formatRt(rt)

Arguments

Details

This function is used to convert retention times. Conversion is seconds to/from the more human friendly format "mm:sec".

Value

A vector of same length as rt .

Author

Laurent Gatto lg390@cam.ac.uk

Examples

formatRt(1524)
formatRt("25:24")

Return a variable name

Description

Return the name of variable varname in call match_call .

Usage

getVariableName(match_call, varname)

Arguments

Value

A character with the name of the variable passed as parameter varname in parent close of match_call .

Author

Laurent Gatto

Examples

a <- 1
f <- function(x, y)
MSnbase:::getVariableName(match.call(), "x")
f(x = a)
f(y = a)

Amino acids

Description

Returns a data.frame of amino acid properties: AA , ResidueMass , Abbrev3 , ImmoniumIonMass , Name , Hydrophobicity , Hydrophilicity , SideChainMass , pK1 , pK2 and pI .

Usage

get.amino.acids()

Value

A data.frame

Author

Laurent Gatto

Examples

get.amino.acids()

Atomic mass.

Description

Returns a double of used atomic mass.

Usage

get.atomic.mass()

Value

A named double .

Author

Sebastian Gibb

Examples

get.atomic.mass()

Returns the matching column names of indices.

Description

Given a text spread sheet f and a pattern to be matched to its header (first line in the file), the function returns the matching columns names or indices of the corresponding data.frame .

Usage

grepEcols(f, pattern, ..., n = 1)
getEcols(f, ..., n = 1)

Arguments

Details

The function starts by reading the first line of the file (or connection) f with readLines , then splits it according to the optional ... arguments (it is important to correctly specify strsplit 's split character vector here) and then matches pattern to the individual column names using grep .

Similarly, getEcols can be used to explore the column names and decide for the appropriate pattern value.

These functions are useful to check the parameters to be provided to readMSnSet2 .

Value

Depending on value , the matching column names of indices. In case of getEcols , a character of column names.

Seealso

readMSnSet2

Author

Laurent Gatto

Checks if raw data files have any spectra or chromatograms

Description

Helper functions to check whether raw files contain spectra or chromatograms.

Usage

hasSpectra(files)
hasChromatograms(files)

Arguments

Value

A logical(n) where n == length(x) with TRUE if that files contains at least one spectrum, FALSE otherwise.

Author

Laurent Gatto

Examples

f <- msdata::proteomics(full.names = TRUE)[1:2]
hasSpectra(f)
hasChromatograms(f)

iPQF: iTRAQ (and TMT) Protein Quantification based on Features

Description

The iPQF spectra-to-protein summarisation method integrates peptide spectra characteristics and quantitative values for protein quantitation estimation. Spectra features, such as charge state, sequence length, identification score and others, contain valuable information concerning quantification accuracy. The iPQF algorithm assigns weights to spectra according to their overall feature reliability and computes a weighted mean to estimate protein quantities. See also combineFeatures for a more general overview of feature aggregation and examples.

Usage

iPQF(object, groupBy, low.support.filter = FALSE, ratio.calc = "sum",
  method.combine = FALSE, feature.weight = c(7, 6, 4, 3, 2, 1, 5)^2)

Arguments

Value

A matrix with estimated protein ratios.

Author

Martina Fischer

References

iPQF: a new peptide-to-protein summarization method using peptide spectra characteristics to improve protein quantification. Fischer M, Renard BY. Bioinformatics. 2016 Apr 1;32(7):1040-7. doi:10.1093/bioinformatics/btv675. Epub 2015 Nov 20. PubMed PMID:26589272.

Examples

data(msnset2)
head(exprs(msnset2))
prot <- combineFeatures(msnset2,
groupBy = fData(msnset2)$accession,
method = "iPQF")
head(exprs(prot))

iTRAQ 4-plex set

Description

This instance of class " corresponds to the iTRAQ 4-plex set, i.e the 114, 115, 116 and 117 isobaric tags. In the iTRAQ5 data set, an unfragmented tag, i.e reporter and attached isobaric tag, is also included at MZ 145. These objects are used to plot the reporter ions of interest in an MSMS spectra (see " ) as well as for quantification (see quantify ).

Usage

iTRAQ4
iTRAQ5
iTRAQ8
iTRAQ9

Seealso

TMT6 .

References

Ross PL, Huang YN, Marchese JN, Williamson B, Parker K, Hattan S, Khainovski N, Pillai S, Dey S, Daniels S, Purkayastha S, Juhasz P, Martin S, Bartlet-Jones M, He F, Jacobson A, Pappin DJ. "Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents." Mol Cell Proteomics , 2004 Dec;3(12):1154-69. Epub 2004 Sep 22. PubMed PMID: 15385600.

Examples

iTRAQ4
iTRAQ4[1:2]

newReporter <- new("ReporterIons",
description="an example",
name="my reporter ions",
reporterNames=c("myrep1","myrep2"),
mz=c(121,122),
col=c("red","blue"),
width=0.05)
newReporter

NA heatmap visualisation for 2 groups

Description

Produces a heatmap after reordring rows and columsn to highlight missing value patterns.

Usage

imageNA2(object, pcol, Rowv, Colv = TRUE, useGroupMean = FALSE,
  plot = TRUE, ...)

Arguments

Value

Used for its side effect of plotting. Invisibly returns Rovw and Colv.

Author

Laurent Gatto, Samuel Wieczorek and Thomas Burger

Examples

library("pRolocdata")
library("pRoloc")
data(dunkley2006)
pcol <- ifelse(dunkley2006$fraction <= 5, "A", "B")
nax <- makeNaData(dunkley2006, pNA = 0.10)
exprs(nax)[sample(nrow(nax), 30), pcol == "A"] <- NA
exprs(nax)[sample(nrow(nax), 50), pcol == "B"] <- NA
MSnbase:::imageNA2(nax, pcol)
MSnbase:::imageNA2(nax, pcol, useGroupMean = TRUE)
MSnbase:::imageNA2(nax, pcol, Colv = FALSE, useGroupMean = FALSE)
MSnbase:::imageNA2(nax, pcol, Colv = FALSE, useGroupMean = TRUE)

Quantitative proteomics data imputation

Description

The impute method performs data imputation on an MSnSet instance using a variety of methods (see below). The imputation and the parameters are logged into the processingData(object) slot.

Users should proceed with care when imputing data and take precautions to assure that the imputation produce valid results, in particular with naive imputations such as replacing missing values with 0.

Details

There are two types of mechanisms resulting in missing values in LC/MSMS experiments.

• list("Missing values resulting from absence of detection of a ", " feature, despite ions being present at detectable concentrations. ", " For example in the case of ion suppression or as a result from the ", " stochastic, data-dependent nature of the MS acquisition ", " method. These missing value are expected to be randomly ", " distributed in the data and are defined as missing at random (MAR) ", " or missing completely at random (MCAR). ")

• list("Biologically relevant missing values resulting from the ", " absence of the low abundance of ions (below the limit of detection ", " of the instrument). These missing values are not expected to be ", " randomly distributed in the data and are defined as missing not at ", " random (MNAR). ")

MNAR features should ideally be imputed with a left-censor method, such as QRILC below. Conversely, it is recommended to use host deck methods such nearest neighbours, Bayesian missing value imputation or maximum likelihood methods when values are missing at random.

Currently, the following imputation methods are available:

list(" ", " ", " ", list(list("MLE"), list("Maximum likelihood-based imputation method using the EM ", " algorithm. Implemented in the ", list("norm::imp.norm"), " function. See ", " ", list(list("imp.norm")), " for details and additional ", " parameters. Note that here, ", list("..."), " are passed to the ", " ", list(list("em.norm")), " function, rather to the actual ", " imputation function ", list("imp.norm"), ". ")), " ", " ", " ", list(list("bpca"), list(

"Bayesian missing value imputation are available, as

", " implemented in the and ", list("pcaMethods::pca"), " functions. See ", " ", list(list("pca")), " for details and additional ", " parameters. ")), " ", " ", " ", list(list("knn"), list("Nearest neighbour averaging, as implemented in the ", " ", list("impute::impute.knn"), " function. See ", " ", list(list("impute.knn")), " for details and additional ", " parameters. ")), " ", " ", " ", list(list("QRILC"),

list("A missing data imputation method that performs the

", " imputation of left-censored missing data using random draws from a ", " truncated distribution with parameters estimated using quantile ", " regression. Implemented in the ", list("imputeLCMD::impute.QRILC"), " ", " function. See ", list(list("impute.QRILC")), " for details ", " and additional parameters. ")), " ", " ", " ", list(list("MinDet"), list("Performs the imputation of left-censored missing data ",

"      using a deterministic minimal value approach. Considering a

", " expression data with ", list("n"), " samples and ", list("p"), " features, for ", " each sample, the missing entries are replaced with a minimal value ", " observed in that sample. The minimal value observed is estimated as ", " being the q-th quantile (default ", list("q = 0.01"), ") of the observed ", " values in that sample. Implemented in the ", " ", list("imputeLCMD::impute.MinDet"),

" function. See

", " ", list(list("impute.MinDet")), " for details and additional ", " parameters. ")), " ", " ", " ", list(list("MinProb"), list("Performs the imputation of left-censored missing data ", " by random draws from a Gaussian distribution centred to a minimal ", " value. Considering an expression data matrix with ", list("n"), " samples ", " and ", list("p"), " features, for each sample, the mean value of the ", " Gaussian distribution is set to a minimal observed value in that ",

"      sample. The minimal value observed is estimated as being the q-th

", " quantile (default ", list("q = 0.01"), ") of the observed values in that ", " sample. The standard deviation is estimated as the median of the ", " feature standard deviations. Note that when estimating the ", " standard deviation of the Gaussian distribution, only the ", " peptides/proteins which present more than 50% recorded values are ", " considered. Implemented in the ", list(

    "imputeLCMD::impute.MinProb"), "

", " function. See ", list(list("impute.MinProb")), " for details ", " and additional parameters. ")), " ", " ", " ", list(list("min"), list("Replaces the missing values by the smallest non-missing ", " value in the data. ")), " ", " ", " ", list(list("zero"), list("Replaces the missing values by 0.")), " ", " ", " ", list(list("mixed"), list("A mixed imputation applying two methods (to be defined ", " by the user as ",

list("mar"), " for values missing at random and

", " ", list("mnar"), " for values missing not at random, see example) on two ", " M[C]AR/MNAR subsets of the data (as defined by the user by a ", " ", list("randna"), " logical, of length equal to ", list("nrow(object)"), "). ")), " ", " ", " ", list(list("nbavg"), list("Average neighbour imputation for fractions collected ", " along a fractionation/separation gradient, such as sub-cellular ", " fractions. The method assumes that the fraction are ordered along ",

"      the gradient and is invalid otherwise.

", " ", " Continuous sets ", list("NA"), " value at the beginning and the end of ", " the quantitation vectors are set to the lowest observed value in ", " the data or to a user defined value passed as argument ", list("k"), ". ", " Them, when a missing value is flanked by two non-missing ", " neighbouring values, it is imputed by the mean of its direct ", " neighbours. A stretch of 2 or more missing values will not be ",

"      imputed. See the example below.

", " ", " ")), " ", " ", " ", list(list("none"), list("No imputation is performed and the missing values are ", " left untouched. Implemented in case one wants to only impute value ", " missing at random or not at random with the ", list("mixed"), " method.")), " ", " ", " ")

The naset list("MSnSet") is an real quantitative data where quantitative values have been replaced by NA s. See script/naset.R for details.

Author

Laurent Gatto and Samuel Wieczorek

References

Olga Troyanskaya, Michael Cantor, Gavin Sherlock, Pat Brown, Trevor Hastie, Robert Tibshirani, David Botstein and Russ B. Altman, Missing value estimation methods for DNA microarrays Bioinformatics (2001) 17 (6): 520-525.

Oba et al., A Bayesian missing value estimation method for gene expression profile data, Bioinformatics (2003) 19 (16): 2088-2096.

Cosmin Lazar (2015). imputeLCMD: A collection of methods for left-censored missing data imputation. R package version 2.0. http://CRAN.R-project.org/package=imputeLCMD .

Lazar C, Gatto L, Ferro M, Bruley C, Burger T. Accounting for the Multiple Natures of Missing Values in Label-Free Quantitative Proteomics Data Sets to Compare Imputation Strategies. J Proteome Res. 2016 Apr 1;15(4):1116-25. doi: 10.1021/acs.jproteome.5b00981. PubMed PMID: 26906401.

Examples

data(naset)
## table of missing values along the rows
table(fData(naset)$nNA)
## table of missing values along the columns
pData(naset)$nNA

## non-random missing values
notna <- which(!fData(naset)$randna)
length(notna)
notna

impute(naset, method = "min")

if (require("imputeLCMD")) {
impute(naset, method = "QRILC")
impute(naset, method = "MinDet")
}

if (require("norm"))
impute(naset, method = "MLE")

impute(naset, "mixed",
randna = fData(naset)$randna,
mar = "knn", mnar = "QRILC")

## neighbour averaging

x <- naset[1:4, 1:6]
exprs(x)[1, 1] <- NA ## min value
exprs(x)[2, 3] <- NA ## average
exprs(x)[3, 1:2] <- NA ## min value and average
## 4th row: no imputation
exprs(x)

exprs(impute(x, "nbavg"))

Get mode from mzML data file

Description

The function extracts the mode (profile or centroided) from the raw mass spectrometry file by parsing the mzML file directly. If the object x stems from any other type of file, NA s are returned.

Usage

isCentroidedFromFile(x)

Arguments

Details

This function is much faster than isCentroided() , which estimates mode from the data, but is limited to data stemming from mzML files which are still available in their original location (and accessed with fileNames(x) ).

Value

A named logical vector of the same length as x .

Author

Laurent Gatto

Examples

library("msdata")
f <- proteomics(full.names = TRUE,
pattern = "TMT_Erwinia_1uLSike_Top10HCD_isol2_45stepped_60min_01.mzML.gz")
x <- readMSData(f, mode = "onDisk")
table(isCentroidedFromFile(x), msLevel(x))

Example MSnExp and MSnSet data sets

Description

itraqdata is and example data sets is an iTRAQ 4-plex experiment that has been run on an Orbitrap Velos instrument. It includes identification data in the feature data slot obtain from the Mascot search engine. It is a subset of an spike-in experiment where proteins have spiked in an list("Erwinia") background, as described in

Karp et al. (2010), list("Addressing accuracy and precision issues in ", " iTRAQ quantitation") , Mol Cell Proteomics. 2010 Sep;9(9):1885-97. Epub 2010 Apr 10. (PMID 20382981).

The spiked-in proteins in itradata are BSA and ENO and are present in relative abundances 1, 2.5, 5, 10 and 10, 5, 2.5, 1 in the 114, 115, 116 and 117 reporter tags.

The msnset object is produced by running the quantify method on the itraqdata experimental data, as detailed in the quantify example. This example data set is used in the MSnbase-demo vignette, available with vignette("MSnbase-demo", .

The msnset2 object is another example iTRAQ4 data that is used to demonstrate features of the package, in particular the iPQF feature aggregation method, described in iPQF . It corresponds to 11 proteins with spectra measurements from the original data set described by Breitwieser et al. (2011) list("General ", " statistical modeling of data from protein relative expression isobaric ", " tags") . J. Proteome Res., 10, 2758-2766.

Usage

itraqdata

Examples

data(itraqdata)
itraqdata

## created by
## msnset <- quantify(itraqdata, method = "trap", reporters = iTRAQ4)
data(msnset)
msnset

data(msnset2)
msnset2

Tests equality of list elements class

Description

Compares equality of all members of a list.

Usage

listOf(x, class, valid = TRUE)

Arguments

Value

TRUE is all elements of x inherit from class .

Author

Laurent Gatto

Examples

listOf(list(), "foo")
listOf(list("a", "b"), "character")
listOf(list("a", 1), "character")

Convert to camel case by replacing dots by captial letters

Description

Convert a vector of characters to camel case by replacing dots by captial letters.

Usage

makeCamelCase(x, prefix)

Arguments

Value

A character of same length as x .

Author

Laurent Gatto

Examples

nms <- c("aa.foo", "ab.bar")
makeCamelCase(nms)
makeCamelCase(nms, prefix = "x")

Create a data with missing values

Description

These functions take an instance of class " and sets randomly selected values to NA .

Usage

makeNaData(object, nNA, pNA, exclude)
makeNaData2(object, nRows, nNAs, exclude)
whichNA(x)

Arguments

Details

makeNaData randomly selects a number nNA (or a proportion pNA ) of cells in the expression matrix to be set to NA .

makeNaData2 will select length(nRows) sets of rows from object , each with nRows[i] rows respectively. The first set will be assigned nNAs[1] missing values, the second nNAs[2] , ... As opposed to makeNaData , this permits to control the number of NAs per rows.

The whichNA can be used to extract the indices of the missing values, as illustrated in the example.

Value

An instance of class MSnSet , as object , but with the appropriate number/proportion of missing values. The returned object has an additional feature meta-data columns, nNA

Author

Laurent Gatto

Examples

## Example 1
library(pRolocdata)
data(dunkley2006)
sum(is.na(dunkley2006))
dunkleyNA <- makeNaData(dunkley2006, nNA = 150)
processingData(dunkleyNA)
sum(is.na(dunkleyNA))
table(fData(dunkleyNA)$nNA)
naIdx <- whichNA(dunkleyNA)
head(naIdx)
## Example 2
dunkleyNA <- makeNaData(dunkley2006, nNA = 150, exclude = 1:10)
processingData(dunkleyNA)
table(fData(dunkleyNA)$nNA[1:10])
table(fData(dunkleyNA)$nNA)
## Example 3
nr <- rep(10, 5)
na <- 1:5
x <- makeNaData2(dunkley2006[1:100, 1:5],
nRows = nr,
nNAs = na)
processingData(x)
(res <- table(fData(x)$nNA))
stopifnot(as.numeric(names(res)[-1]) ==  na)
stopifnot(res[-1] ==  nr)
## Example 3
nr2 <- c(5, 12, 11, 8)
na2 <- c(3, 8, 1, 4)
x2 <- makeNaData2(dunkley2006[1:100, 1:10],
nRows = nr2,
nNAs = na2)
processingData(x2)
(res2 <- table(fData(x2)$nNA))
stopifnot(as.numeric(names(res2)[-1]) ==  sort(na2))
stopifnot(res2[-1] ==  nr2[order(na2)])
## Example 5
nr3 <- c(5, 12, 11, 8)
na3 <- c(3, 8, 1, 3)
x3 <- makeNaData2(dunkley2006[1:100, 1:10],
nRows = nr3,
nNAs = na3)
processingData(x3)
(res3 <- table(fData(x3)$nNA))

Combine a list of spectra to a single spectrum

Description

Combine peaks from several spectra into a single spectrum. Intensity and m/z values from the input spectra are aggregated into a single peak if the difference between their m/z values is smaller than mzd or smaller than ppm of their m/z. While mzd can be used to group mass peaks with a single fixed value, ppm allows a m/z dependent mass peak grouping. Intensity values of grouped mass peaks are aggregated with the intensityFun , m/z values by the mean, or intensity weighted mean if weighted = TRUE .

Usage

meanMzInts(x, ..., intensityFun = base::mean, weighted = FALSE,
  main = 1L, mzd, ppm = 0, timeDomain = FALSE, unionPeaks = TRUE)

Arguments

Details

For general merging of spectra, the mzd and/or ppm should be manually specified based on the precision of the MS instrument. Peaks from spectra with a difference in their m/z being smaller than mzd or smaller than ppm of their m/z are grouped into the same final peak.

Some details for the combination of consecutive spectra of an LCMS run:

The m/z values of the same ion in consecutive scans (spectra) of a LCMS run will not be identical. Assuming that this random variation is much smaller than the resolution of the MS instrument (i.e. the difference between m/z values within each single spectrum), m/z value groups are defined across the spectra and those containing m/z values of the main spectrum are retained. The maximum allowed difference between m/z values for the same ion is estimated as in estimateMzScattering() . Alternatively it is possible to define this maximal m/z difference with the mzd parameter. All m/z values with a difference smaller than this value are combined to a m/z group. Intensities and m/z values falling within each of these m/z groups are aggregated using the intensity_fun and mz_fun , respectively. It is highly likely that all QTOF profile data is collected with a timing circuit that collects data points with regular intervals of time that are then later converted into m/z values based on the relationship t = k * sqrt(m/z) . The m/z scale is thus non-linear and the m/z scattering (which is in fact caused by small variations in the time circuit) will thus be different in the lower and upper m/z scale. m/z-intensity pairs from consecutive scans to be combined are therefore defined by default on the square root of the m/z values. With timeDomain = FALSE , the actual m/z values will be used.

Value

Spectrum with m/z and intensity values representing the aggregated values across the provided spectra. The returned spectrum contains the union of all peaks from all spectra (if unionPeaks = TRUE ), or the same number of m/z and intensity pairs than the spectrum with index main in x (if unionPeaks = FALSE . All other spectrum data (such as retention time etc) is taken from the main spectrum.

Seealso

estimateMzScattering() for a function to estimate m/z scattering in consecutive scans.

estimateMzResolution() for a function estimating the m/z resolution of a spectrum.

combineSpectraMovingWindow() for the function to combine consecutive spectra of an MSnExp object using a moving window approach.

Other spectra combination functions: consensusSpectrum

Note

This allows e.g. to combine profile-mode spectra of consecutive scans into the values for the main spectrum. This can improve centroiding of profile-mode data by increasing the signal-to-noise ratio and is used in the combineSpectraMovingWindow() function.

Author

Johannes Rainer, Sigurdur Smarason

Examples

library(MSnbase)
## Create 3 example profile-mode spectra with a resolution of 0.1 and small
## random variations on these m/z values on consecutive scans.
set.seed(123)
mzs <- seq(1, 20, 0.1)
ints1 <- abs(rnorm(length(mzs), 10))
ints1[11:20] <- c(15, 30, 90, 200, 500, 300, 100, 70, 40, 20) # add peak
ints2 <- abs(rnorm(length(mzs), 10))
ints2[11:20] <- c(15, 30, 60, 120, 300, 200, 90, 60, 30, 23)
ints3 <- abs(rnorm(length(mzs), 10))
ints3[11:20] <- c(13, 20, 50, 100, 200, 100, 80, 40, 30, 20)

## Create the spectra.
sp1 <- new("Spectrum1", mz = mzs + rnorm(length(mzs), sd = 0.01),
intensity = ints1)
sp2 <- new("Spectrum1", mz = mzs + rnorm(length(mzs), sd = 0.01),
intensity = ints2)
sp3 <- new("Spectrum1", mz = mzs + rnorm(length(mzs), sd = 0.009),
intensity = ints3)

## Combine the spectra
sp_agg <- meanMzInts(list(sp1, sp2, sp3))

## Plot the spectra before and after combining
par(mfrow = c(2, 1), mar = c(4.3, 4, 1, 1))
plot(mz(sp1), intensity(sp1), xlim = range(mzs[5:25]), type = "h", col = "red")
points(mz(sp2), intensity(sp2), type = "h", col = "green")
points(mz(sp3), intensity(sp3), type = "h", col = "blue")
plot(mz(sp_agg), intensity(sp_agg), xlim = range(mzs[5:25]), type = "h",
col = "black")

Documenting missing data visualisation

Description

There is a need for adequate handling of missing value impuation in quantitative proteomics. Before developing a framework to handle missing data imputation optimally, we propose a set of visualisation tools. This document serves as an internal notebook for current progress and ideas that will eventually materialise in exported functionality in the MSnbase package.

Details

The explore the structure of missing values, we propose to

1. Explore missing values in the frame of the experimental design. The imageNA2 function offers such a simple visualisation. It is currently limited to 2-group designs/comparisons. In case of time course experiments or sub-cellular fractionation along a density gradient, we propose to split the time/gradient into 2 groups (early/late, top/bottom) as a first approximation.

2. Explore the proportion of missing values in each group.

3. Explore the total and group-wise feature intensity distributions.

The existing plotNA function illustrates the completeness/missingness of the data.

Seealso

plotNA , imageNA2 .

Author

Laurent Gatto lg390@cam.ac.uk, Samuel Wieczorek and Thomas Burger

Examples

## Other suggestions
library("pRolocdata")
library("pRoloc")
data(dunkley2006)
set.seed(1)
nax <- makeNaData(dunkley2006, pNA = 0.10)
pcol <- factor(ifelse(dunkley2006$fraction <= 5, "A", "B"))
sel1 <- pcol == "A"

## missing values in each sample
barplot(colSums(is.na(nax)), col = pcol)


## table of missing values in proteins
par(mfrow = c(3, 1))
barplot(table(rowSums(is.na(nax))), main = "All")
barplot(table(rowSums(is.na(nax)[sel1,])), main = "Group A")
barplot(table(rowSums(is.na(nax)[!sel1,])), main = "Group B")


fData(nax)$nNA1 <- rowSums(is.na(nax)[, sel1])
fData(nax)$nNA2 <- rowSums(is.na(nax)[, !sel1])
fData(nax)$nNA <- rowSums(is.na(nax))
o <- MSnbase:::imageNA2(nax, pcol)

plot((fData(nax)$nNA1 - fData(nax)$nNA2)[o], type = "l")
grid()

plot(sort(fData(nax)$nNA1 - fData(nax)$nNA2), type = "l")
grid()


o2 <- order(fData(nax)$nNA1 - fData(nax)$nNA2)
MSnbase:::imageNA2(nax, pcol, Rowv=o2)

layout(matrix(c(rep(1, 10), rep(2, 5)), nc = 3))
MSnbase:::imageNA2(nax, pcol, Rowv=o2)
plot((fData(nax)$nNA1 - fData(nax)$nNA)[o2], type = "l", col = "red",
ylim = c(-9, 9), ylab = "")
lines((fData(nax)$nNA - fData(nax)$nNA2)[o2], col = "steelblue")
lines((fData(nax)$nNA1 - fData(nax)$nNA2)[o2], type = "l",
lwd = 2)

Coerce identification data to a data.frame

Description

A function to convert the identification data contained in an mzRident object to a data.frame . Each row represents a scan, which can however be repeated several times if the PSM matches multiple proteins and/or contains two or more modifications. To reduce the data.frame so that rows/scans are unique and use semicolon-separated values to combine information pertaining a scan, use reduce .

Arguments

Details

See also the Tandem MS identification data section in the MSnbase-demo vignette.

Value

A data.frame

Author

Laurent Gatto

Examples

## find path to a mzIdentML file
identFile <- dir(system.file(package = "MSnbase", dir = "extdata"),
full.name = TRUE, pattern = "dummyiTRAQ.mzid")
library("mzR")
x <- openIDfile(identFile)
x
as(x, "data.frame")

How many features in a group?

Description

This function computes the number of features in the group defined by the feature variable fcol and appends this information in the feature data of object .

Usage

nFeatures(object, fcol)

Arguments

Value

An updated MSnSet with a new feature variable fcol.nFeatures .

Author

Laurent Gatto

Examples

library(pRolocdata)
data("hyperLOPIT2015ms3r1psm")
hyperLOPIT2015ms3r1psm <- nFeatures(hyperLOPIT2015ms3r1psm,
"Protein.Group.Accessions")
i <- c("Protein.Group.Accessions", "Protein.Group.Accessions.nFeatures")
fData(hyperLOPIT2015ms3r1psm)[1:10, i]

Count the number of quantitfied features.

Description

This function counts the number of quantified features, i.e non NA quantitation values, for each group of features for all the samples in an " object. The group of features are defined by a feature variable names, i.e the name of a column of fData(object) .

Usage

nQuants(x, groupBy)

Arguments

Details

This function is typically used after topN and before combineFeatures , when the summerising function is sum , or any function that does not normalise to the number of features aggregated. In the former case, sums of features might be the result of 0 (if no feature was quantified) to n (if all topN 's n features were quantified) features, and one might want to rescale the sums based on the number of non-NA features effectively summed.

Value

A matrix of dimensions length(levels(groupBy)) by ncol(x)

A matrix of dimensions length(levels(factor(fData(object)[, fcol]))) by ncol(object) of integers.

Author

Laurent Gatto lg390@cam.ac.uk, Sebastian Gibb mail@sebastiangibb.de

Examples

data(msnset)
n <- 2
msnset <- topN(msnset, groupBy = fData(msnset)$ProteinAccession, n)
m <- nQuants(msnset, groupBy = fData(msnset)$ProteinAccession)
msnset2 <- combineFeatures(msnset,
groupBy = fData(msnset)$ProteinAccession,
method = sum)
stopifnot(dim(n) == dim(msnset2))
head(exprs(msnset2))
head(exprs(msnset2) * (n/m))

Overview of missing value

Description

Visualise missing values as a heatmap and barplots along the samples and features.

Usage

naplot(object, verbose = isMSnbaseVerbose(), reorderRows = TRUE,
  reorderColumns = TRUE, ...)

Arguments

Value

Used for its side effect. Invisibly returns NULL

Author

Laurent Gatto

Examples

data(naset)
naplot(naset)

Navigate an MSnExp object

Description

Navigate an MSnExp object by moving to the next or previous spectrum.

Usage

navMS(i, object, msLevel, nav = c("nextMS", "prevMS"), ...)
nextMS(...)
prevMS(...)

Arguments

Value

An object of class Spectrum1 or Spectrum2 , depending on the value of msLevel or NULL , of no spectrum is found.

Author

Laurent Gatto

Examples

f <- msdata::proteomics(full.names = TRUE, pattern = "MS3")
x <- readMSData(f, centroided. = c(FALSE, TRUE, FALSE), mode = "onDisk")
(sp <- which(msLevel(x) == 3)[2]) ## 2nd MS3 spectrum
x[[sp]] ## curent MS3
MSnbase:::nextMS(sp, x) ## next MS3
MSnbase:::prevMS(sp, x) ## prev MS3
MSnbase:::prevMS(sp, x, 2L) ## prev MS2
MSnbase:::prevMS(sp, x, 1L) ## prev MS1

Combine peptides into proteins.

Description

This function combines peptides into their proteins by normalising the intensity values to a reference run/sample for each protein.

Usage

normToReference(x, group, reference = .referenceFractionValues(x = x,
  group = group))

Arguments

Details

This function is not intented to be used directly (that's why it is not exported via NAMESPACE ). Instead the user should use combineFeatures .

The algorithm is described in Nikolovski et al., briefly it works as follows:

• Find reference run (column) for each protein (grouped rows). We use the run (column) with the lowest number of NA . If multiple candidates are available we use the one with the highest intensity. This step is skipped if the user use his own reference vector.

• For each protein (grouped rows) and each run (column):

• Find peptides (grouped rows) shared by the current run (column) and the reference run (column).

• Sum the shared peptides (grouped rows) for the current run (column) and the reference run (column).

• The ratio of the shared peptides (grouped rows) of the current run (column) and the reference run (column) is the new intensity for the current protein for the current run.

Value

a matrix with one row per protein.

Seealso

combineFeatures

Author

Sebastian Gibb mail@sebastiangibb.de, Pavel Shliaha

References

Nikolovski N, Shliaha PV, Gatto L, Dupree P, Lilley KS. Label-free protein quantification for plant Golgi protein localization and abundance. Plant Physiol. 2014 Oct;166(2):1033-43. DOI: 10.1104/pp.114.245589. PubMed PMID: 25122472.

Examples

library("MSnbase")
data(msnset)

# choose the reference run automatically
combineFeatures(msnset, groupBy=fData(msnset)$ProteinAccession)

# use a user-given reference
combineFeatures(msnset, groupBy=fData(msnset)$ProteinAccession,
reference=rep(2, 55))

Normalisation of MSnExp , MSnSet and Spectrum objects

Description

The normalise method (also available as normalize ) performs basic normalisation on spectra intensities of single spectra ( " or " objects), whole experiments ( " objects) or quantified expression data ( " objects).

Raw spectra and experiments are normalised using max or sum only. For MSMS spectra could be normalised to their precursor additionally. Each peak intensity is divided by the highest intensity in the spectrum, the sum of intensities or the intensity of the precursor. These methods aim at facilitating relative peaks heights between different spectra.

The method parameter for " can be one of sum , max , quantiles , center.mean , center.median , .median , quantiles.robust or vsn . For sum and max , each feature's reporter intensity is divided by the maximum or the sum respectively. These two methods are applied along the features (rows).

center.mean and center.median translate the respective sample (column) intensities according to the column mean or median. diff.median translates all samples (columns) so that they all match the grand median. Using quantiles or quantiles.robust applies (robust) quantile normalisation, as implemented in normalize.quantiles and normalize.quantiles.robust of the preprocessCore package. vsn uses the vsn2 function from the vsn package. Note that the latter also glog-transforms the intensities. See respective manuals for more details and function arguments.

A scale method, mimicking the base scale method exists for " instances. See ?base:: for details.

Arguments

Examples

## quantifying full experiment
data(msnset)
msnset.nrm <- normalise(msnset, "quantiles")
msnset.nrm

Non-parametric coefficient of variation

Description

Calculates a non-parametric version of the coefficient of variation where the standard deviation is replaced by the median absolute deviations (see mad for details) and divided by the absolute value of the mean.

Usage

npcv(x, na.rm = TRUE)

Arguments

Details

Note that the mad of a single value is 0 (as opposed to NA for the standard deviation, see example below).

Value

A numeric .

Author

Laurent Gatto

Examples

set.seed(1)
npcv(rnorm(10))
replicate(10, npcv(rnorm(10)))
npcv(1)
mad(1)
sd(1)

Class to Contain Raw Mass-Spectrometry Assays and Experimental Metadata

Description

Container for high-throughput mass-spectrometry assays and experimental metadata. This class is based on Biobase's " virtual class, with the notable exception that 'assayData' slot is an environment contain objects of class " .

Seealso

" for an instantiatable application of pSet .

Author

Laurent Gatto lg390@cam.ac.uk

References

The " class, on which pSet is based.

Examples

showClass("pSet")

Peak Detection for 'MSnExp' or 'Spectrum' instances

Description

This method performs a peak picking on individual spectra ( Spectrum instances) or whole experiments ( MSnExp instances) to create centroided spectra. For noisy spectra there are currently two different noise estimators available, the Median Absolute Deviation ( method = "MAD" ) and Friedman's Super Smoother ( method = "SuperSmoother" ), as implemented in the MALDIquant::detectPeaks and MALDIquant::estimateNoise functions respectively.

The method supports also to optionally refine the m/z value of the identified centroids by considering data points that belong (most likely) to the same mass peak. The m/z value is calculated as an intensity weighted average of the m/z values within the peak region. How the peak region is defined depends on the method chosen:

refineMz = "kNeighbors" : m/z values (and their respective intensities) of the 2 * k closest signals to the centroid are used in the intensity weighted average calculation. The number of neighboring signals can be defined with the argument k .

refineMz = "descendPeak" : the peak region is defined by descending from the identified centroid/peak on both sides until the measured signal increases again. Within this defined region all measurements with an intensity of at least signalPercentage of the centroid's intensity are used to calculate the refined m/z. By default the descend is stopped when the first signal that is equal or larger than the last observed one is encountered. Setting stopAtTwo = TRUE , two consecutively increasing signals are required.

By default ( refineMz = "none" , simply the m/z of the largest signal (the identified centroid) is reported. See below for examples.

Seealso

clean , removePeaks smooth , estimateNoise and trimMz for other spectra processing methods.

Author

Sebastian Gibb mail@sebastiangibb.de with contributions from Johannes Rainer.

References

S. Gibb and K. Strimmer. 2012. MALDIquant: a versatile R package for the analysis of mass spectrometry data. Bioinformatics 28: 2270-2271. http://strimmerlab.org/software/maldiquant/

Examples

sp1 <- new("Spectrum1",
intensity = c(1:6, 5:1),
mz = 1:11,
centroided = FALSE)
sp2 <- pickPeaks(sp1)
intensity(sp2)

data(itraqdata)
itraqdata2 <- pickPeaks(itraqdata)
processingData(itraqdata2)


## Examples for refineMz:
ints <- c(5, 3, 2, 3, 1, 2, 4, 6, 8, 11, 4, 7, 5, 2, 1, 0, 1, 0, 1, 1, 1, 0)
mzs <- 1:length(ints)
sp1 <- new("Spectrum1", intensity = ints, mz = mzs, centroided = FALSE)
plot(mz(sp1), intensity(sp1), type = "h")

## Without m/z refinement:
sp2 <- pickPeaks(sp1)
points(mz(sp2), intensity(sp2), col = "darkgrey")
## Using k = 1, closest signals
sp3 <- pickPeaks(sp1, refineMz = "kNeighbors", k = 1)
points(mz(sp3), intensity(sp3), col = "green", type = "h")

## Using descendPeak requiring at least 50% or the centroid's intensity
sp4 <- pickPeaks(sp1, refineMz = "descendPeak", signalPercentage = 50)
points(mz(sp4), intensity(sp4), col = "red", type = "h")

The 'plot2d' method for 'MSnExp' quality assessment

Description

These methods plot the retention time vs. precursor MZ for the whole " experiment. Individual dots will be colour-coded to describe individual spectra's peaks count, total ion count, precursor charge (MS2 only) or file of origin.

The methods make use the ggplot2 system. An object of class 'ggplot' is returned invisibly.

Arguments

Seealso

The plotDensity and plotMzDelta methods for other QC plots.

Author

Laurent Gatto lg390@cam.ac.uk

Examples

itraqdata
plot2d(itraqdata,z="ionCount")
plot2d(itraqdata,z="peaks.count")
plot2d(itraqdata,z="charge")

The 'plotDensity' method for 'MSnExp' quality assessment

Description

These methods plot the distribution of several parameters of interest for the different precursor charges for " experiment.

The methods make use the ggplot2 system. An object of class 'ggplot' is returned invisibly.

Arguments

Seealso

The plot2d and plotDensity methods for other QC plots.

Author

Laurent Gatto lg390@cam.ac.uk

Examples

itraqdata
plotDensity(itraqdata,z="ionCount")
plotDensity(itraqdata,z="peaks.count")
plotDensity(itraqdata,z="precursor.mz")

The delta m/z plot

Description

The m/z delta plot illustrates the suitability of MS2 spectra for identification by plotting the m/z differences of the most intense peaks. The resulting histogram should optimally shown outstanding bars at amino acid residu masses. The plots have been described in Foster et al 2011.

Only a certain percentage of most intense MS2 peaks are taken into account to use the most significant signal. Default value is 10% (see percentage argument). The difference between peaks is then computed for all individual spectra and their distribution is plotted as a histogram where single bars represent 1 m/z differences. Delta m/z between 40 and 200 are plotted by default, to encompass the residue masses of all amino acids and several common contaminants, although this can be changes with the xlim argument.

In addition to the processing described above, isobaric reporter tag peaks (see the reporters argument) and the precursor peak (see the precMz argument) can also be removed from the MS2 spectrum, to avoid interence with the fragment peaks.

Note that figures in Foster et al 2011 have been produced and optimised for centroided data. Application of the plot as is for data in profile mode has not been tested thoroughly, although the example below suggest that it might work.

The methods make use the ggplot2 system. An object of class ggplot is returned invisibly.

Most of the code for plotMzDelta has kindly been contributed by Guangchuang Yu.

Arguments

Seealso

The plotDensity and plot2d methods for other QC plots.

Author

Laurent Gatto lg390@cam.ac.uk and Guangchuang Yu

References

Foster JM, Degroeve S, Gatto L, Visser M, Wang R, Griss J, Apweiler R, Martens L. "A posteriori quality control for the curation and reuse of public proteomics data." Proteomics , 2011 Jun;11(11):2182-94. doi:10.1002/pmic.201000602. Epub 2011 May 2. PMID: 21538885

Examples

mzdplot <- plotMzDelta(itraqdata,
subset = 0.5,
reporters = iTRAQ4,
verbose = FALSE, plot = FALSE)
## let's retrieve peptide sequence information
## and get a table of amino acids
peps <- as.character(fData(itraqdata)$PeptideSequence)
aas <- unlist(strsplit(peps,""))
## table of aas
table(aas)
## mzDelta plot
print(mzdplot)

Exploring missing data in 'MSnSet' instances

Description

These methods produce plots that illustrate missing data.

is.na returns the expression matrix of it MSnSet argument as a matrix of logicals referring whether the corresponding cells are NA or not. It is generally used in conjunction with table and image (see example below).

The plotNA method produces plots that illustrate missing data. The completeness of the full dataset or a set of proteins (ordered by increasing NA content along the x axis) is represented. The methods make use the ggplot2 system. An object of class 'ggplot' is returned invisibly.

Seealso

See also the filterNA method to filter out features with a specified proportion if missing values.

Author

Laurent Gatto lg390@cam.ac.uk

Examples

data(msnset)
exprs(msnset)[sample(prod(dim(msnset)), 120)] <- NA

head(is.na(msnset))
table(is.na(msnset))
image(msnset)

plotNA(msnset, pNA = 1/4)

Plotting a 'Spectrum' vs another 'Spectrum' object.

Description

These method plot mass spectra MZ values against the intensities as line plots. The first spectrum is plotted in the upper panel and the other in upside down in the lower panel. Common peaks are drawn in a slightly darker colour. If a peptide sequence is provided it automatically calculates and labels the fragments.

Arguments

Seealso

More spectrum plotting available in plot.Spectrum .

More details about fragment calculation: calculateFragments .

Author

Sebastian Gibb mail@sebastiangibb.de

Examples

## find path to a mzXML file
file <- dir(system.file(package = "MSnbase", dir = "extdata"),
full.name = TRUE, pattern = "mzXML$")

## create basic MSnExp
msexp <- readMSData(file, centroided.=FALSE)

## centroid them
msexp <- pickPeaks(msexp)

## plot the first against the second spectrum
plot(msexp[[1]], msexp[[2]])

## add sequence information
plot(msexp[[1]], msexp[[2]], sequences=c("VESITARHGEVLQLRPK",
"IDGQWVTHQWLKK"))


itraqdata2 <- pickPeaks(itraqdata)
(k <- which(fData(itraqdata2)[, "PeptideSequence"] == "TAGIQIVADDLTVTNPK"))
mzk <- precursorMz(itraqdata2)[k]
zk <- precursorCharge(itraqdata2)[k]
mzk * zk
plot(itraqdata2[[k[1]]], itraqdata2[[k[2]]])

Plotting 'MSnExp' and 'Spectrum' object(s)

Description

These methods provide the functionality to plot mass spectrometry data provided as MSnExp , OnDiskMSnExp or Spectrum objects. Most functions plot mass spectra M/Z values against intensities.

Full spectra (using the full parameter) or specific peaks of interest can be plotted using the reporters parameter. If reporters are specified and full is set to 'TRUE', a sub-figure of the reporter ions is inlaid inside the full spectrum.

If an " is provided as argument, all the spectra are aligned vertically. Experiments can be subset to extract spectra of interest using the [ operator or extractPrecSpectra methods.

Most methods make use the ggplot2 system in which case an object of class 'ggplot' is returned invisibly.

If a single " and a "character" representing a valid peptide sequence are passed as argument, the expected fragement ions are calculated and matched/annotated on the spectum plot.

Arguments

Seealso

calculateFragments to calculate ions produced by fragmentation and plot.Spectrum.Spectrum to plot and compare 2 spectra and their shared peaks.

Chromatogram for plotting of chromatographic data.

Author

Laurent Gatto lg390@cam.ac.uk, Johannes Rainer and Sebastian Gibb

Examples

data(itraqdata)
## plotting experiments
plot(itraqdata[1:2], reporters = iTRAQ4)
plot(itraqdata[1:2], full = TRUE)
## plotting spectra
plot(itraqdata[[1]],reporters = iTRAQ4, full = TRUE)

itraqdata2 <- pickPeaks(itraqdata)
i <- 14
s <- as.character(fData(itraqdata2)[i, "PeptideSequence"])
plot(itraqdata2[[i]], s, main = s)

## Load profile-mode LC-MS files
library(msdata)
od <- readMSData(dir(system.file("sciex", package = "msdata"),
full.names = TRUE), mode = "onDisk")
## Restrict the MS data to signal for serine
serine <- filterMz(filterRt(od, rt = c(175, 190)), mz = c(106.04, 106.06))
plot(serine, type = "XIC")

## Same plot but using heat.colors, rectangles and no point border
plot(serine, type = "XIC", pch = 22, colramp = heat.colors, col = NA)

Number of precursor selection events

Description

precSelection computes the number of selection events each precursor ions has undergone in an tandem MS experiment. This will be a function of amount of peptide loaded, chromatography efficiency, exclusion time,... and is useful when optimising and experimental setup. This function returns a named integer vector or length equal to the number of unique precursor MZ values in the original experiment. See n parameter to set the number of MZ significant decimals.

precSelectionTable is a wrapper around precSelection and returns a table with the number of single, 2-fold, ... selection events.

Usage

precSelection(object,n)

Arguments

Value

A named integer in case of precSelection and a table for precSelectionTable .

Author

Laurent Gatto lg390@cam.ac.uk

Examples

precSelection(itraqdata)
precSelection(itraqdata,n=2)
precSelectionTable(itraqdata)
## only single selection event in this reduced exeriment

Performs reporter ions purity correction

Description

Manufacturers sometimes provide purity correction values indicating the percentages of each reporter ion that have masses differing by +/- n Da from the nominal reporter ion mass due to isotopic variants. This correction is generally applied after reporter peaks quantitation.

Purity correction here is applied using solve from the base package using the purity correction values as coefficient of the linear system and the reporter quantities as the right-hand side of the linear system. 'NA' values are ignored and negative intensities after correction are also set to 'NA'.

A more elaborated purity correction method is described in Shadforth et al. , i-Tracker: for quantitative proteomics using iTRAQ. BMC Genomics. 2005 Oct 20;6:145. (PMID 16242023).

Function makeImpuritiesMatrix(x, filename, edit = TRUE) helps the user to create such a matrix. The function can be used in two ways. If given an integer x , it is used as the dimension of the square matrix (i.e the number of reporter ions). For TMT6-plex and iTRAQ4-plex, default values taken from manufacturer's certification sheets are used as templates, but batch specific values should be used whenever possible. Alternatively, the filename of a csv spreadsheet can be provided. The sheet should define the correction factors as illustrated below (including reporter names in the first column and header row) and the corresponding correction matrix is calculated. Examples of such csv files are available in the package's extdata directory. Use dir(system.file("extdata", package = "MSnbase"), pattern = to locate them. If edit = TRUE , the the matrix can be edited before it is returned.

Arguments

|impurities | A square 'matrix' of dim equal to ncol(object) defining the correction coefficients to be applied. The reporter ions should be ordered along the columns and the relative percentages along the rows. As an example, below is the correction factors as provided in an ABI iTRAQ 4-plex certificate of analysis: list(list("lrrrr"), list(" ", " reporter ", list(), " % of -2 ", list(), " % of -1 ", list(), " % of +1 ", list(), " % of +2 ", list(), " ", " 114 ", list(), " 0.0 ", list(), " 1.0 ", list(), " 5.9 ", list(), " 0.2 ", list(), " ", " 115 ", list(), " 0.0 ", list(), " 2.0 ", list(), " 5.6 ", list(), " 0.1 ", list(), " ", " 116 ", list(), " 0.0 ", list(), " 3.0 ", list(), " 4.5 ", list(), " 0.1 ", list(), " ", " 117 ", list(), " 0.1 ", list(), " 4.0 ", list(), " 3.5 ", |

list(), " 0.1 ", list(), "

", " ")) The impurity table will be list(list("rrrr"), list(" ", " 0.929 ", list(), " 0.059 ", list(), " 0.002 ", list(), " 0.000 ", list(), " ", " 0.020 ", list(), " 0.923 ", list(), " 0.056 ", list(), " 0.001 ", list(), " ", " 0.000 ", list(), " 0.030 ", list(), " 0.924 ", list(), " 0.045 ", list(), " ", " 0.000 ", list(), " 0.001 ", list(), " 0.040 ", list(), " 0.923 ", list(), " ", " ")) where, the diagonal is computed as 100 - sum of rows of the original table and subsequent cells are directly filled in. Similarly, for TMT 6-plex tags, we observe list(list("lrrrrrr"), list(" ", " reporter ", list(), " % of -3 ", list(), " % of -2 ", list(), " % of -1 ", list(), " % of +1 % ", list(), " % of +2 ", list(), " % of +3 ", list(), " ", " 126 ", list(), " 0 ", list(), " 0 ", list(), " 0 ", list(), " 6.1 ", list(), " 0 ", list(), " 0 ", list(), " ", " 127 ", list(), " 0 ", list(), " 0 ", list(), " 0.5 ", list(), " 6.7 ", list(), " 0 ", list(), " 0 ", list(), " ", " 128 ", list(), " 0 ", list(), " 0 ", list(), " 1.1 ", list(),

" 4.2 ", list(), " 0 ", list(), " 0 ", list(), "

", " 129 ", list(), " 0 ", list(), " 0 ", list(), " 1.7 ", list(), " 4.1 ", list(), " 0 ", list(), " 0 ", list(), " ", " 130 ", list(), " 0 ", list(), " 0 ", list(), " 1.6 ", list(), " 2.1 ", list(), " 0 ", list(), " 0 ", list(), " ", " 131 ", list(), " 0 ", list(), " 0.2 ", list(), " 3.2 ", list(), " 2.8 ", list(), " 0 ", list(), " 0 ", list(), " ", " ")) and obtain the following impurity correction matrix list(list("rrrrrr"), list(" ", " 0.939 ", list(), " 0.061 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " ", " 0.005 ", list(), " 0.928 ", list(), " 0.067 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " ", " 0.000 ", list(), " 0.011 ", list(), " 0.947 ", list(), " 0.042 ", list(), " 0.000 ", list(), " 0.000 ", list(), " ", " 0.000 ", list(), " 0.000 ", list(), " 0.017 ", list(), " 0.942 ", list(), " 0.041 ", list(),

" 0.000 ", list(), "

", " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.016 ", list(), " 0.963 ", list(), " 0.021 ", list(), " ", " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.002 ", list(), " 0.032 ", list(), " 0.938 ", list(), " ", " ")) For iTRAQ 8-plex, given the following correction factors (to make such a matrix square, if suffices to add -4, -3, +3 and +4 columns filled with zeros): list(list("lllll"), list(" ", " TAG ", list(), " -2 ", list(), " -1 ", list(), " +1 ", list(), " +2 ", list(), " ", " 113 ", list(), " 0 ", list(), " 2.5 ", list(), " 3 ", list(), " 0.1 ", list(), " ", " 114 ", list(), " 0 ", list(), " 1 ", list(), " 5.9 ", list(), " 0.2 ", list(), " ", " 115 ", list(), " 0 ", list(), " 2 ", list(), " 5.6 ", list(), " 0.1 ", list(), " ", " 116 ", list(), " 0 ", list(), " 3 ", list(), " 4.5 ", list(), " 0.1 ", list(), " ",

"      117 ", list(), " 0.1 ", list(), " 4 ", list(), "  3.5 ", list(), " 0.1 ", list(), "

", " 118 ", list(), " 0.1 ", list(), " 2 ", list(), " 3 ", list(), " 0.1 ", list(), " ", " 119 ", list(), " 0.1 ", list(), " 2 ", list(), " 4 ", list(), " 0.1 ", list(), " ", " 121 ", list(), " 0.1 ", list(), " 2 ", list(), " 3 ", list(), " 0.1 ", list(), " ", " ")) we calculate the impurity correction matrix shown below list(list("lrrrrrrrr"), list(" ", " ", list(), " 113 ", list(), " 114 ", list(), " 115 ", list(), " 116 ", list(), " 117 ", list(), " 118 ", list(), " 119 ", list(), " 121 ", list(), " ", " % reporter 113 ", list(), " 0.944 ", list(), " 0.030 ", list(), " 0.001 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " ", " % reporter 114 ", list(), " 0.010 ", list(), " 0.929 ", list(), " 0.059 ", list(), " 0.002 ",

list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), "

", " % reporter 115 ", list(), " 0.000 ", list(), " 0.020 ", list(), " 0.923 ", list(), " 0.056 ", list(), " 0.001 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " ", " % reporter 116 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.030 ", list(), " 0.924 ", list(), " 0.045 ", list(), " 0.001 ", list(), " 0.000 ", list(), " 0.000 ", list(), " ", " % reporter 117 ",

list(), " 0.000 ", list(), " 0.000 ", list(), " 0.001 ", list(), " 0.040 ", list(), " 0.923 ", list(), " 0.035 ", list(), " 0.001 ", list(), " 0.000 ", list(), "

", " % reporter 118 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.001 ", list(), " 0.020 ", list(), " 0.948 ", list(), " 0.030 ", list(), " 0.001 ", list(), " ", " % reporter 119 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.001 ", list(), " 0.020 ", list(),

" 0.938 ", list(), " 0.040 ", list(), "

", " % reporter 121 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.001 ", list(), " 0.020 ", list(), " 0.948 ", list(), " ", " ")) Finally, for a TMT 10-plex impurity matrix (for example lot RH239932 ) list(list("llllllllll"), list(" ", " . ", list(), " -2 ", list(), " -1 ", list(), " 1 ", list(), " 2 ", list(), " ", " 126 ", list(), " 0.0 ", list(), " 0.0 ", list(), " 5.0 (127C) ", list(), " 0.0 (128C) ", list(), " ", " 127N ", list(), " 0.0 ", list(), " 0.2 ", list(), " 5.8 (128N) ", list(), " 0.0 (129N) ", list(), " ", " 127C ", list(), " 0.0 ", list(), " 0.3 (126) ", list(), " 4.8 (128C) ", list(),

" 0.0 (129C) ", list(), "

", " 128N ", list(), " 0.0 ", list(), " 0.4 (127N) ", list(), " 4.1 (129N) ", list(), " 0.0 (130N) ", list(), " ", " 128C ", list(), " 0.0 (126) ", list(), " 0.6 (127C) ", list(), " 3.0 (129C) ", list(), " 0.0 (130C) ", list(), " ", " 129N ", list(), " 0.0 (127N) ", list(), " 0.8 (128N) ", list(), " 3.5 (130N) ", list(), " 0.0 (131) ", list(), " ", " 129C ", list(), " 0.0 (127C) ", list(), " 1.4 (128C) ", list(), " 2.4 (130C) ", list(),

" 0.0 ", list(), "

", " 130N ", list(), " 0.1 (128N) ", list(), " 1.5 (129N) ", list(), " 2.4 (131) ", list(), " 3.2 ", list(), " ", " 130C ", list(), " 0.0 (128C) ", list(), " 1.7 (129C) ", list(), " 1.8 ", list(), " 0.0 ", list(), " ", " 131 ", list(), " 0.2 (129N) ", list(), " 2.0 (130N) ", list(), " 2.2 ", list(), " 0.0 ", list(), " ", " ")) (Note that a previous example, taken from lot PB199188A , contained a typo.) the impurity correction matrix is list(list("lllllllllll"), list(" ", " ", " . ", list(), " 126 ", list(), " 127N ", list(), " 127C ", list(), " 128N ", list(), " 128C ", list(), " 129N ", list(), " 129C ", list(), " 130N ", list(), " 130C ", list(), " 131 ", list(), " ", " % reporter 126 ", list(), " 0.950 ", list(), " 0.000 ", list(), " 0.050 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " ", " % reporter 127N ",

list(), " 0.000 ", list(), " 0.940 ", list(), " 0.000 ", list(), " 0.058 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), "

", " % reporter 127C ", list(), " 0.003 ", list(), " 0.000 ", list(), " 0.949 ", list(), " 0.000 ", list(), " 0.048 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " ", " % reporter 128N ", list(), " 0.000 ", list(), " 0.004 ",

list(), " 0.000 ", list(), " 0.955 ", list(), " 0.000 ", list(), " 0.041 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), "

", " % reporter 128C ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.006 ", list(), " 0.000 ", list(), " 0.964 ", list(), " 0.000 ", list(), " 0.030 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " ", " % reporter 129N ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.008 ",

list(), " 0.000 ", list(), " 0.957 ", list(), " 0.000 ", list(), " 0.035 ", list(), " 0.000 ", list(), " 0.000 ", list(), "

", " % reporter 129C ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.014 ", list(), " 0.000 ", list(), " 0.962 ", list(), " 0.000 ", list(), " 0.024 ", list(), " 0.000 ", list(), " ", " % reporter 130N ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.001 ", list(), " 0.000 ", list(), " 0.015 ",

list(), " 0.000 ", list(), " 0.928 ", list(), " 0.000 ", list(), " 0.024 ", list(), "

", " % reporter 130C ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.017 ", list(), " 0.000 ", list(), " 0.965 ", list(), " 0.000 ", list(), " ", " % reporter 131 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.000 ", list(), " 0.002 ", list(), " 0.000 ", list(), " 0.020 ",

list(), " 0.000 ", list(), " 0.956 ", list(), "

", " ")) These examples are provided as defaults impurity correction matrices in makeImpuritiesMatrix .

Examples

## quantifying full experiment
data(msnset)
impurities <- matrix(c(0.929,0.059,0.002,0.000,
0.020,0.923,0.056,0.001,
0.000,0.030,0.924,0.045,
0.000,0.001,0.040,0.923),
nrow=4, byrow = TRUE)
## or, using makeImpuritiesMatrix()
impurities <- makeImpuritiesMatrix(4)
msnset.crct <- purityCorrect(msnset, impurities)
head(exprs(msnset))
head(exprs(msnset.crct))
processingData(msnset.crct)

## default impurity matrix for iTRAQ 8-plex
makeImpuritiesMatrix(8, edit = FALSE)

## default impurity matrix for TMT 10-plex
makeImpuritiesMatrix(10, edit = FALSE)

Quantifies 'MSnExp' and 'Spectrum' objects

Description

This method quantifies individual " objects or full " experiments. Current, MS2-level isobar tagging using iTRAQ and TMT (or any arbitrary peaks of interest, see " ) and MS2-level label-free quantitation (spectral counting, spectral index or spectral abundance factor) are available.

Isobaric tag peaks of single spectra or complete experiments can be quantified using appropriate methods . Label-free quantitation is available only for MSnExp experiments.

Since version 1.13.5, parallel quantitation is supported by the BiocParallel package and controlled by the BPPARAM argument.

Arguments

Details

" define specific MZ at which peaks are expected and a window around that MZ value. A peak of interest is searched for in that window. Since version 1.1.2, warnings are not thrown anymore in case no data is found in that region or if the peak extends outside the window. This can be checked manually after quantitation, by inspecting the quantitation data (using the exprs accessor) for NA values or by comaring the lowerMz and upperMz columns in the " qual slot against the respective expected mz(reporters) +/- width(reporters) .

Once the range of the curve is found, quantification is performed. If no data points are found in the expected region, NA is returned for the reporter peak MZ.

Note that for label-free, spectra that have not been identified (the corresponding fields in the feature data are populated with NA values) or that have been uniquely assigned to a protein (the nprot feature data is greater that 1) are removed prior to quantitation. The latter does not apply for method = "count" but can be applied manually with removeMultipleAssignment .

Author

Laurent Gatto lg390@cam.ac.uk and Sebastian Gibb mail@sebastiangibb.de

References

For details about the spectral index (SI), see Griffin NM, Yu J, Long F, Oh P, Shore S, Li Y, Koziol JA, Schnitzer JE. list("Label-free, ", " normalized quantification of complex mass spectrometry data for ", " proteomic analysis") . Nat Biotechnol. 2010 Jan;28(1):83-9. doi: 10.1038/nbt.1592. PMID: 20010810; PubMed Central PMCID: PMC2805705.

For details about the spectra abundance factor, see Paoletti AC, Parmely TJ, Tomomori-Sato C, Sato S, Zhu D, Conaway RC, Conaway JW, Florens L, Washburn MP. list("Quantitative proteomic analysis of ", " distinct mammalian Mediator complexes using normalized spectral ", " abundance factors") . PNAS. 2006 Dec 12;103(50):18928-33. PMID: 17138671; PubMed Central PMCID: PMC1672612.

Examples

## Quantifying a full experiment using iTRAQ4-plex tagging
data(itraqdata)
msnset <- quantify(itraqdata, method = "trap", reporters = iTRAQ4)
msnset

## specifying a custom parallel framework
## bp <- MulticoreParam(2L) # on Linux/OSX
## bp <- SnowParam(2L) # on Windows
## quantify(itraqdata[1:10], method = "trap", iTRAQ4, BPPARAM = bp)

## Checking for non-quantified peaks
sum(is.na(exprs(msnset)))

## Quantifying a single spectrum
qty <- quantify(itraqdata[[1]], method = "trap", iTRAQ4[1])
qty$peakQuant
qty$curveStats


## Label-free quantitation
## Raw (mzXML) and identification (mzid) files
quantFile <- dir(system.file(package = "MSnbase", dir = "extdata"),
full.name = TRUE, pattern = "mzXML$")
identFile <- dir(system.file(package = "MSnbase", dir = "extdata"),
full.name = TRUE, pattern = "dummyiTRAQ.mzid")

msexp <- readMSData(quantFile)
msexp <- addIdentificationData(msexp, identFile)
fData(msexp)$DatabaseAccess

si <- quantify(msexp, method = "SIn")
processingData(si)
exprs(si)

saf <- quantify(msexp, method = "NSAF")
processingData(saf)
exprs(saf)

Imports mass-spectrometry raw data files as 'MSnExp' instances.

Description

Reads as set of XML-based mass-spectrometry data files and generates an MSnExp object. This function uses the functionality provided by the mzR package to access data and meta data in mzData , mzXML and mzML .

Usage

readMSData(files, pdata = NULL, msLevel. = NULL,
  verbose = isMSnbaseVerbose(), centroided. = NA, smoothed. = NA,
  cache. = 1L, mode = c("inMemory", "onDisk"))

Arguments

Details

When using the inMemory mode, the whole MS data is read from file and kept in memory as Spectrum objects within the MSnExp 'es assayData slot.

To reduce the memory footpring especially for large MS1 data sets it is also possible to read only selected information from the MS files and fetch the actual spectrum data (i.e. the M/Z and intensity values) only on demand from the original data files. This can be achieved by setting mode = "onDisk" . The function returns then an OnDiskMSnExp object instead of a MSnExp object.

Value

An MSnExp object for inMemory mode and a OnDiskMSnExp object for onDisk mode.

Seealso

readMgfData() to read mgf peak lists.

Note

readMSData uses normalizePath to replace relative with absolute file paths.

Author

Laurent Gatto

Examples

file <- dir(system.file(package = "MSnbase", dir = "extdata"),
full.name = TRUE,
pattern = "mzXML$")
mem <- readMSData(file, mode = "inMemory")
mem
dsk <- readMSData(file, mode = "onDisk")
dsk

Read 'MSnSet'

Description

This function reads data files to generate an MSnSet instance. It is a wrapper around Biobase 's readExpressionSet function with an additional featureDataFile parameter to include feature data. See also readExpressionSet for more details. readMSnSet2 is a simple version that takes a single text spreadsheet as input and extracts the expression data and feature meta-data to create and MSnSet .

Note that when using readMSnSet2 , one should not set rownames as additional argument to defined feature names. It is ignored and used to set fnames if not provided otherwise.

Usage

readMSnSet(exprsFile,
           phenoDataFile,
           featureDataFile,
           experimentDataFile,
           notesFile,
           path, annotation,
           exprsArgs = list(sep = sep, header = header, row.names = row.names, quote = quote, ...),
           phenoDataArgs = list(sep = sep, header = header, row.names = row.names, quote = quote, stringsAsFactors = stringsAsFactors, ...),
           featureDataArgs = list(sep = sep, header = header, row.names = row.names, quote = quote, stringsAsFactors = stringsAsFactors, ...),
           experimentDataArgs = list(sep = sep, header = header, row.names = row.names, quote = quote, stringsAsFactors = stringsAsFactors, ...),
           sep = "  ",
           header = TRUE,
           quote = "",
           stringsAsFactors = FALSE,
           row.names = 1L,
           widget = getOption("BioC")$Base$use.widgets, ...)
readMSnSet2(file, ecol, fnames, ...)

Arguments

Value

An instance of the MSnSet class.

Seealso

The grepEcols and getEcols helper functions to identify the ecol values. The MSnbase-io vignette illustrates these functions in detail. It can be accessed with vignette("MSnbase-io") .

Author

Laurent Gatto lg390@cam.ac.uk

Examples

exprsFile <- "path_to_intensity_file.csv"
fdatafile <- "path_to_featuredata_file.csv"
pdatafile <- "path_to_sampledata_file.csv"
## Read ExpressionSet with appropriate parameters
res <- readMSnSet(exprsFile, pdataFile, fdataFile, sep = "  ", header=TRUE)

library("pRolocdata")
f0 <- dir(system.file("extdata", package = "pRolocdata"),
full.names = TRUE,
pattern = "Dunkley2006")
basename(f0)
res <- readMSnSet2(f0, ecol = 5:20)
res
head(exprs(res)) ## columns 5 to 20
head(fData(res)) ## other columns

Import mgf files as 'MSnExp' instances.

Description

Reads a mgf file and generates an " object.

Usage

readMgfData(filename, pdata = NULL, centroided = TRUE, smoothed = FALSE,
verbose = isMSnbaseVerbose(), cache = 1)

Arguments

Details

Note that when reading an mgf file, the original order of the spectra is lost. Thus, if the data was originally written to mgf from an MSnExp object using writeMgfData , although the feature names will be identical, the spectra are not as a result of the reordering. See example below.

Value

An instance of

Seealso

writeMgfData method to write the content of " or " objects to mgf files. Raw data files can also be read with the readMSData function.

Author

Guangchuang Yu guangchuangyu@gmail.com and Laurent Gatto lg390@cam.ac.uk

Examples

data(itraqdata)
writeMgfData(itraqdata, con="itraqdata.mgf", COM="MSnbase itraqdata")
itraqdata2 <- readMgfData("itraqdata.mgf")
## note that the order of the spectra is altered
## and precision of some values (precursorMz for instance)
match(signif(precursorMz(itraqdata2),4),signif(precursorMz(itraqdata),4))
## [1]  1 10 11 12 13 14 15 16 17 18 ...
## ... but all the precursors are there
all.equal(sort(precursorMz(itraqdata2)),
sort(precursorMz(itraqdata)),
check.attributes=FALSE,
tolerance=10e-5)
## is TRUE
all.equal(as.data.frame(itraqdata2[[1]]),as.data.frame(itraqdata[[1]]))
## is TRUE
all.equal(as.data.frame(itraqdata2[[3]]),as.data.frame(itraqdata[[11]]))
## is TRUE
f <- dir(system.file(package="MSnbase",dir="extdata"),
full.name=TRUE,
pattern="test.mgf")
(x <- readMgfData(f))
x[[2]]
precursorMz(x[[2]])
precursorIntensity(x[[2]])
precursorMz(x[[1]])
precursorIntensity(x[[1]]) ## was not in test.mgf
scanIndex(x)

Import peptide-spectrum matches

Description

Reads as set of mzId files containing PSMs an generates a data.frame .

Usage

readMzIdData(files)

Arguments

Details

This function uses the functionality provided by the mzR package to access data in the mzId files. An object of class mzRident can also be coerced to a data.frame using as(, "data.frame") .

Value

A data.frame containing the PSMs stored in the mzId files.

Seealso

filterIdentificationDataFrame() to filter out unreliable PSMs.

Author

Laurent Gatto

Examples

idf <- "TMT_Erwinia_1uLSike_Top10HCD_isol2_45stepped_60min_01-20141210.mzid"
f <- msdata::ident(full.names = TRUE, pattern = idf)
basename(f)
readMzIdData(f)

Read an 'mzTab' file

Description

This function can be used to create an " by reading and parsing an mzTab file. The metadata section is always used to populate the MSnSet 's experimentData()@other$mzTab slot.

Usage

readMzTabData(file, what = c("PRT", "PEP", "PSM"), version = c("1.0",
  "0.9"), verbose = isMSnbaseVerbose())

Arguments

Value

An instance of class MSnSet .

Seealso

See MzTab and MSnSetList for details about the inners of readMzTabData .

Author

Laurent Gatto

Examples

testfile <- "https://raw.githubusercontent.com/HUPO-PSI/mzTab/master/examples/1_0-Proteomics-Release/PRIDE_Exp_Complete_Ac_16649.xml-mztab.txt"
prot <- readMzTabData(testfile, "PRT")
prot
head(fData(prot))
head(exprs(prot))
psms <- readMzTabData(testfile, "PSM")
psms
head(fData(psms))

Read an 'mzTab' file

Description

This function can be used to create a " by reading and parsing an mzTab file. The metadata section is always used to populate the MSnSet 's experimentData slot.

Usage

readMzTabData_v0.9(file, what = c("PRT", "PEP"),
  verbose = isMSnbaseVerbose())

Arguments

Value

An instance of class MSnSet .

Seealso

writeMzTabData to save an " as an mzTab file.

Author

Laurent Gatto

Examples

testfile <- "https://raw.githubusercontent.com/HUPO-PSI/mzTab/master/legacy/jmztab-1.0/examples/mztab_itraq_example.txt"
prot <- readMzTabData_v0.9(testfile, "PRT")
prot
pep <- readMzTabData_v0.9(testfile, "PEP")
pep

Read SRM/MRM chromatographic data

Description

The readSRMData function reads MRM/SRM data from provided mzML files and returns the results as a Chromatograms() object.

Usage

readSRMData(files, pdata = NULL)

Arguments

Details

readSRMData supports reading chromatogram entries from mzML files. If multiple files are provided the same precursor and product m/z for SRM/MRM chromatograms are expected across files. The number of columns of the resulting Chromatograms() object corresponds to the number of files. Each row in the Chromatograms() object is supposed to contain chromatograms with same polarity, precursor and product m/z. If chromatograms with redundant polarity, precursor and product m/z values are found, they are placed into multiple consecutive rows in the Chromatograms() object.

Value

A Chromatograms() object. See details above for more information.

Note

readSRMData reads only SRM/MRM chromatogram data, i.e. chromatogram data from mzML files with precursorIsolationWindowTargetMZ and productIsolationWindowTargetMZ attributes. Total ion chromatogram data is hence not extracted.

The number of features and hence rows of the resulting Chromatograms object depends on the total list of unique precursor and product m/z isolation windows found across all input files. In cases in which not each file has chromatgraphic data for the same polarity, precursor and product m/z, an empty Chromatogram() object is reported for the specific precursor and product m/z combination of the respective file (and a warning is thrown).

Author

Johannes Rainer

Examples

## Read an example MRM/SRM data
library(msdata)
fl <- proteomics(full.names = TRUE, pattern = "MRM")

## Read the data
mrm <- readSRMData(fl)

## The data is represented as a Chromatograms object, each column
## containing the data from one input file
mrm

## Access the polarity for each chromatogram (row)
polarity(mrm)

## Access the precursor m/z. The result is returned as a matrix with
## columns representing the minimum and maximum m/z (will be identical in
## most cases).
precursorMz(mrm)

## Access the product m/z.
productMz(mrm)

## Plot one chromatogram
plot(mrm[1, ])

Reduce a data.frame

Description

Reduce a data.frame so that the (primary) key column contains only unique entries and other columns pertaining to that entry are combined into semicolon-separated values into a single row/observation.

Usage

list(list("reduce"), list("data.frame"))(x, key, sep = ";")

Arguments

Details

An important side-effect of reducing a data.frame is that all columns other than the key are converted to characters when they are collapsed to a semi-column separated value (even if only one value is present) as soon as one observation of transformed.

Value

A reduced data.frame .

Author

Laurent Gatto

Examples

dfr <- data.frame(A = c(1, 1, 2),
B = c("x", "x", "z"),
C = LETTERS[1:3])
dfr
dfr2 <- reduce(dfr, key = "A")
dfr2
## column A used as key is still num
str(dfr2)
dfr3 <- reduce(dfr, key = "B")
dfr3
## A is converted to chr; B remains factor
str(dfr3)
dfr4 <- data.frame(A = 1:3,
B = LETTERS[1:3],
C = c(TRUE, FALSE, NA))
## No effect of reducing, column classes are maintained
str(reduce(dfr4, key = "B"))

Removes non-identified features

Description

The method removes non-identifed features in MSnExp and MSnSet instances using relevant information from the feaureData slot of a user-provide filtering vector of logicals.

Seealso

MSnExp and MSnSet .

Author

Laurent Gatto lg390@cam.ac.uk

Examples

quantFile <- dir(system.file(package = "MSnbase", dir = "extdata"),
full.name = TRUE, pattern = "mzXML$")
identFile <- dir(system.file(package = "MSnbase", dir = "extdata"),
full.name = TRUE, pattern = "dummyiTRAQ.mzid")
msexp <- readMSData(quantFile)
msexp <- addIdentificationData(msexp, identFile)
fData(msexp)$sequence
length(msexp)

## using default fcol
msexp2 <- removeNoId(msexp)
length(msexp2)
fData(msexp2)$sequence

## using keep
print(fvarLabels(msexp))
(k <- fData(msexp)$'MS.GF.EValue' > 75)
k[is.na(k)] <- FALSE
k
msexp3 <- removeNoId(msexp, keep = k)
length(msexp3)
fData(msexp3)$sequence

Removes low intensity peaks

Description

This method sets low intensity peaks from individual spectra ( Spectrum instances) or whole experiments ( MSnExp instances) to 0. The intensity threshold is set with the t parameter. Default is the "min" character. The threshold is then set as the non-0 minimum intensity found in the spectrum. Any other numeric values is valid. All peaks with maximum intensity smaller or equal to t are set to 0.

If the spectrum is in profile mode, ranges of successive non-0 peaks <= t are set to 0. If the spectrum is centroided, then individual peaks <= t are set to 0. See the example below for an illustration.

Note that the number of peaks is not changed; the peaks below the threshold are set to 0 and the object is not cleanded out (see clean ). An illustrative example is shown below.

Seealso

clean and trimMz for other spectra processing methods.

Author

Laurent Gatto lg390@cam.ac.uk

Examples

int <- c(2, 0, 0, 0, 1, 5, 1, 0, 0, 1, 3, 1, 0, 0, 1, 4, 2, 1)
sp1 <- new("Spectrum2",
intensity = int,
mz = 1:length(int),
centroided = FALSE)
sp2 <- removePeaks(sp1) ## no peaks are removed here
## as min intensity is 1 and
## no peak has a max int <= 1
sp3 <- removePeaks(sp1, 3)
intensity(sp1)
intensity(sp2)
intensity(sp3)

peaksCount(sp1) == peaksCount(sp2)
peaksCount(sp3) <= peaksCount(sp1)

data(itraqdata)
itraqdata2 <- removePeaks(itraqdata, t = 2.5e5)
table(unlist(intensity(itraqdata)) == 0)
table(unlist(intensity(itraqdata2)) == 0)
processingData(itraqdata2)

## difference between centroided and profile peaks

int <- c(104, 57, 32, 33, 118, 76, 38, 39, 52, 140, 52, 88, 394, 71,
408, 94, 2032)
sp <- new("Spectrum2",
intensity = int,
centroided = FALSE,
mz = seq_len(length(int)))

## unchanged, as ranges of peaks <= 500 considered
intensity(removePeaks(sp, 500))
stopifnot(identical(intensity(sp), intensity(removePeaks(sp, 500))))

centroided(sp) <- TRUE
## different!
intensity(removePeaks(sp, 500))

Removes reporter ion tag peaks

Description

This methods sets all the reporter tag ion peaks from one MS2 spectrum or all the MS2 spectra of an experiment to 0. Reporter data is specified using an " instance. The peaks are selected around the expected reporter ion m/z value +/- the reporter width. Optionally, the spectrum/spectra can be cleaned to remove successive 0 intensity data points (see the clean function for details).

Note that this method only works for MS2 spectra or experiments that contain MS2 spectra. It will fail for MS1 spectrum.

Seealso

clean and removePeaks for other spectra processing methods.

Author

Laurent Gatto lg390@cam.ac.uk

Examples

sp1 <- itraqdata[[1]]
sp2 <- removeReporters(sp1,reporters=iTRAQ4)
sel <- mz(sp1) > 114 & mz(sp1) < 114.2
mz(sp1)[sel]
intensity(sp1)[sel]
plot(sp1,full=TRUE,reporters=iTRAQ4)
intensity(sp2)[sel]
plot(sp2,full=TRUE,reporters=iTRAQ4)

Select feature variables of interest

Description

Select feature variables to be retained.

requiredFvarLabels returns a character vector with the required feature data variable names ( fvarLabels , i.e. the column names in the fData data.frame ) for the specified object.

Usage

selectFeatureData(object, graphics = TRUE, fcol)
requiredFvarLabels(x = c("OnDiskMSnExp", "MSnExp", "MSnSet"))

Arguments

Value

For selectFeatureData : updated object containing only selected feature variables.

For requiredFvarLabels : character with the required feature variable names.

Author

Laurent Gatto

Examples

library("pRolocdata")
data(hyperLOPIT2015)
## 5 first feature variables
x <- selectFeatureData(hyperLOPIT2015, fcol = 1:5)
fvarLabels(x)
## select via GUI
x <- selectFeatureData(hyperLOPIT2015)
fvarLabels(x)

## Subset the feature data of an OnDiskMSnExp object to the minimal
## required columns
f <- system.file("microtofq/MM14.mzML", package = "msdata")
od <- readMSData(f, mode = "onDisk")

## what columns do we have?
fvarLabels(od)

## Reduce the feature data data.frame to the required columns only
od <- selectFeatureData(od, fcol = requiredFvarLabels(class(od)))
fvarLabels(od)

Smooths 'MSnExp' or 'Spectrum' instances

Description

This method smooths individual spectra ( Spectrum instances) or whole experiments ( MSnExp instances). Currently, the Savitzky-Golay-Smoothing ( method = "SavitzkyGolay" ) and the Moving-Average-Smoothing ( method = "MovingAverage" ) are available, as implemented in the MALDIquant::smoothIntensity function. Additional methods might be added at a later stage.

Seealso

clean , pickPeaks , removePeaks and trimMz for other spectra processing methods.

Author

Sebastian Gibb mail@sebastiangibb.de

References

A. Savitzky and M. J. Golay. 1964. Smoothing and differentiation of data by simplified least squares procedures. Analytical chemistry, 36(8), 1627-1639.

M. U. Bromba and H. Ziegler. 1981. Application hints for Savitzky-Golay digital smoothing filters. Analytical Chemistry, 53(11), 1583-1586.

S. Gibb and K. Strimmer. 2012. MALDIquant: a versatile R package for the analysis of mass spectrometry data. Bioinformatics 28: 2270-2271. http://strimmerlab.org/software/maldiquant/

Examples

sp1 <- new("Spectrum1",
intensity = c(1:6, 5:1),
mz = 1:11)
sp2 <- smooth(sp1, method = "MovingAverage", halfWindowSize = 2)
intensity(sp2)

data(itraqdata)
itraqdata2 <- smooth(itraqdata,
method = "MovingAverage",
halfWindowSize = 2)
processingData(itraqdata2)

Trims 'MSnExp' or 'Spectrum' instances

Description

This method selects a range of MZ values in a single spectrum ( Spectrum instances) or all the spectra of an experiment ( MSnExp instances). The regions to trim are defined by the range of mz argument, such that MZ values <= min(mz) and MZ values >= max(mz) are trimmed away.

Seealso

removePeaks and clean for other spectra processing methods.

Author

Laurent Gatto lg390@cam.ac.uk

Examples

mz <- 1:100
sp1 <- new("Spectrum2",
mz = mz,
intensity = abs(rnorm(length(mz))))

sp2 <- trimMz(sp1, c(25, 75))
range(mz(sp1))
range(mz(sp2))

data(itraqdata)
itraqdata2 <- filterMz(itraqdata, c(113, 117))
range(mz(itraqdata))
range(mz(itraqdata2))
processingData(itraqdata2)

Update MSnbase objects

Description

Methods for function updateObject for objects from the MSnbase package. See updateObject for details.