TopDownSet-class: The TopDownSet class

In topdownr: Investigation of Fragmentation Conditions in Top-Down Proteomics

Description

The TopDownSet class is a container for a whole top-down proteomics experiment.

Usage

## S4 method for signature 'TopDownSet'
aggregate(x, by = x$Sample, ...)

## S4 method for signature 'TopDownSet,TopDownSet'
combine(x, y)

## S4 method for signature 'TopDownSet'
filterCv(object, threshold, by = object$Sample, ...)

## S4 method for signature 'TopDownSet'
filterInjectionTime(
  object,
  maxDeviation = log2(3),
  keepTopN = 2,
  by = object$Sample,
  ...
)

## S4 method for signature 'TopDownSet'
filterIntensity(object, threshold, relative = TRUE, ...)

## S4 method for signature 'TopDownSet'
filterNonReplicatedFragments(object, minN = 2, by = object$Sample, ...)

## S4 method for signature 'TopDownSet'
normalize(object, method = "TIC", ...)

## S4 method for signature 'TopDownSet,missing'
plot(x, y, ..., verbose = interactive())

## S4 method for signature 'TopDownSet'
show(object)

## S4 method for signature 'TopDownSet'
summary(object, what = c("conditions", "fragments"), ...)

Arguments

x, object	TopDownSet
by	list, grouping variable, in general it refers to technical
y	missing, not used.
threshold	double, threshold variable.
maxDeviation	double, maximal allowed deviation in the log2 injection time in ms in comparison to the median ion injection time.
keepTopN	integer, how many technical replicates should be kept?
relative	logical, if relative is TRUE all fragments with intensity below threshold * max(intensity) per fragment are removed, otherwise all fragments below threshold are removed.
minN	numeric, if less than minN of a fragment are found across technical replicates it is removed.
method	character, normalisation method, currently just "TIC" for Total Ion Current normalisation of the scans/conditions (column-wise normalisation) is supported.
verbose	logical, verbose output?
what	character, specifies whether "conditions" (columns; default) or "fragments" (rows) should be summarized.
...	arguments passed to internal/other methods. replicates (that's why the default is the "Sample" column in colData).

Details

See vignette("analysis", package="topdownr") for a detailed example how to work with TopDownSet objects.

Value

An TopDownSet object.

Methods (by generic)

• aggregate: Aggregate conditions/runs.

  Aggregates conditions/runs (columns) in an TopDownSet object by a user-given value (default is the "Sample" column of colData which has the same value for technical replicates). It combines intensity values and numeric metadata of the grouped conditions/runs (columns) by mean and returns a reduced TopDownSet object.

• combine: Combine TopDownSet objects.

  combine allows to combine two or more TopDownSet objects. Please note that it uses the default sampleColumns to define technical replicates (see readTopDownFiles()).and the default by argument to group ion injection times for the calculation of the median time (see updateMedianInjectionTime()). Both could be modified after combine by calling updateConditionNames() (with modified sampleColumns argument) and updateMedianInjectionTime() (with modified by argument).

• filterCv: Filter by CV.

  Filtering is done by coefficient of variation across technical replicates (defined by the by argument). All fragments below a given threshold are removed. The threshold is the maximal allowed CV in percent (sd/mean * 100 < threshold).

• filterInjectionTime: Filter by ion injection time.

  Filtering is done by maximal allowed deviation and just the technical keepTopN replicates with the lowest deviation from the median ion injection time are kept.

• filterIntensity: Filter by intensity.

  Filtering is done by removing all fragments that are below a given (absolute/relative) intensity threshold.

• filterNonReplicatedFragments: Filter by non-replicated fragments.

  Filtering is done by removing all fragments that don't replicate across technical replicates.

• normalize: Normalise.

  Applies Total Ion Current normalisation to a TopDownSet object. The normalisation ist done per scans/conditions (column-wise normalisation).

• plot: Plotting.

  Plots an TopDownSet object. The function returns a list of ggplot objects (one item per condtion). Use pdf or another non-interactive device to plot the list of ggplot objects (see example section).

• summary: Summary statistics.

  Returns a matrix with some statistics: number of fragments, total/min/first quartile/median/mean/third quartile/maximum of intensity values.

Slots

rowViews

FragmentViews, information about fragments (name, type, sequence, mass, charge), see FragmentViews for details.

colData

S4Vectors::DataFrame, information about the MS2 experiments and the fragmentation conditions.

assay

Matrix::dgCMatrix, intensity values of the fragments. The rows correspond to the fragments and the columns to the condition/run. It just stores values that are different from zero.

files

character, files that were imported.

tolerance

double, tolerance in ppm that were used for matching the experimental mz values to the theoretical fragments.

redundantMatching

character, matching strategies for redundant ion/fragment matches. See redundantIonMatch and redundantFragmentMatch in readTopDownFiles() for details.

processing

character, log messages.

Coercion

'as(object, "MSnSet"): Coerce an TopDownSet object into an MSnbase::MSnSet object.

'as(object, "NCBSet"): Coerce an TopDownSet object into an NCBSet object.

Author(s)

Sebastian Gibb [e-mail]

See Also

• FragmentViews for the row view interface.

• Matrix::dgCMatrix for technical details about the intensity storage.

• ?vignette("analysis", package="topdownr") for a full documented example of an analysis using topdownr.

Examples

## Example data
data(tds, package="topdownr")

tds

head(summary(tds))

# Accessing slots
rowViews(tds)
colData(tds)
head(assayData(tds))

# Accessing colData
tds$Mz
tds$FilterString

# Subsetting

# First 100 fragments
tds[1:100]

# All c fragments
tds["c"]

# Just c 152
tds["c152"]

# Condition 1 to 10
tds[, 1:10]

# Filtering
# Filter all intensities that don't have at least 10 % of the highest
# intensity per fragment.
tds <- filterIntensity(tds, threshold=0.1)

# Filter all conditions with a CV above 30 % (across technical replicates)
tds <- filterCv(tds, threshold=30)

# Filter all conditions with a large deviation in injection time
tds <- filterInjectionTime(tds, maxDeviation=log2(3), keepTopN=2)

# Filter all conditions where fragments don't replicate
tds <- filterNonReplicatedFragments(tds)

# Normalise by TIC
tds <- normalize(tds)

# Aggregate technical replicates
tds <- aggregate(tds)

head(summary(tds))

# Coercion
as(tds, "NCBSet")

if (require("MSnbase")) {
    as(tds, "MSnSet")
}
## Not run: 
# plot a single condition
# pseudo-code (replace topdownset with your object)
plot(topdownset[,1])

# plot the whole object
pdf("topdown-spectra.pdf", paper="a4r", width=12)
# pseudo-code (replace topdownset with your object)
plot(topdownset)
dev.off()

## End(Not run)

topdownr documentation built on Nov. 8, 2020, 8:10 p.m.