| Title: | Fingerprinting for Flow Cytometry |
|---|---|
| Description: | Fingerprint generation of flow cytometry data, used to facilitate the application of machine learning and datamining tools for flow cytometry. |
| Authors: | Herb Holyst <[email protected]>, Wade Rogers <[email protected]> |
| Maintainer: | Herb Holyst <[email protected]>, Wade Rogers <[email protected]> |
| License: | Artistic-2.0 |
| Version: | 1.65.0 |
| Built: | 2024-10-30 07:50:21 UTC |
| Source: | https://github.com/bioc/flowFP |
This package is used to explore flow cytometry data through the use of fingerprints. The broad aim of the package is to transform flow cytometric data into a form amenable to algorithmic analysis tools. Thus, it is useful to think of flowFP as an intermediate step between the acquisition of high-throughput flow cytometric data and empirical modeling, machine learning and knowledge discovery.
A fingerprint is a feature vector meant to efficiently represent the multivariate
probability distribution function for a flow cytometry data set. It is produced
by first creating a data-relevant model of a space, and then applying the
model to a dataset, thereby producing fingerprints.
Model creation is done through the flowFPModel constructor which can be
customized via function arguments. After the model is built, it can be applied
to arbitrary flowFrames or
flowSets using the flowFP
constructor. The resulting S4 object implements plotting and summary methods
that allow the user to compare and contrast instances, using the flowFPModel
as a sort of basis representation, akin for example to trigonometric functions in a
Fourier Transform.
This package is closely integrated with flowCore.
You will want to become familiar with it in order to effectively use flowFP.
| Package: | flowFP |
| Type: | Package |
| Version: | 1.1.2 |
| Depends: | R(>= 2.5.0), flowCore, flowViz |
| Collate: | flowFPModel.R flowFP.R |
| Bioinformatics: | Flowcytometry, CellBasedAssays, Clustering, Statistics, Visualization |
| License: | Artistic-2.0 |
| Built: | R 2.8.0; unix |
flowFPModel-class is the fundamental class for the
flowFP package. It represents the multivariate probability distribution
function for a flow cytometry data set. Information is maintained in a number of
slots, which should only be accessed through methods, described below, not
by direct use of the @ operator. For a complete detailed list of slot names and
descriptions look at the flowFPModel-class help page.
flowFP-class extends the flowFPModel and contains
additional slots to record the assignment to and number of events in the bins
of a flowFPModel. Methods are supplied to retrieve and visualize the
the contents of a flowFP.
flowFPPlex-class is a container for a set of congruent flowFP
objects (by congruent, we mean that each flowFP is a description of the same
set of instances). When constructing or appending flowFPs into a plex,
simple error checking is done to ensure each instance in each flowFP is
equivalent. Both the sampleNames and the sampleClasses slots are consulted
for internal consistency. An error is generated if any of the flowFPs
to be joined in a plex contain different sampleNames and/or sampleClasses.
flowFPModel is the constructor for this class.
flowFPModel(fcs, name="Default Model", parameters=NULL, nRecursions="auto",
dequantize=TRUE, sampleSize=NULL)
fcs |
Either an flowFrame or a flowSet used to create the model. |
name |
Name given the model. |
parameters |
Parameters to consider when constructing the model.(e.g. c(1,5)). |
nRecursions |
The number of level of recursive subdivision. |
dequantize |
Setting this value causes a small incremental value to be added |
| to each event starting with 1e-8. This effectively reduces the | |
| number of duplicate values to break ties when binning. | |
sampleSize |
The max number of events to use out of each fcs file in a |
flowFrame or flowSet. |
|
flowFP is the constructor for this class.
flowFP(fcs, model=NULL, sampleClasses=NULL, ...)
fcs |
Either an flowFrame or a flowSet used to create the fingerprint. |
model |
A model created using flowFPModel or NULL in which case |
| a model will be created from the fcs data supplied. | |
sampleClasses |
List of sample class names to be assigned in order to the instances. |
... |
If the model is NULL the parameters are passed through to |
flowFPModel.
|
flowFPPlex is the constructor for this class.
flowFPPlex(fingerprints=NULL)
fingerprints |
Either an single flowFP or a list of flowFPs |
(e.g. fingerprints = c(fp1, fp2)). All of the fingerprints |
|
| must share the same sample names, and class names (or no class names). |
For further information please see the vignette.
Herb Holyst <[email protected]>, Wade Rogers <[email protected]>
M. Roederer, et. al. (2001) Probability Binning Comparison: A Metric for Quantitating Multivariate Distribution Differences, Cytometry 45, 47-55.
W. Rogers et. al. (2008) Cytometric Fingerprinting: Quantitative Characterization of Multivariate Distributions, Cytometry Part A 73, 430-441.
flowCore, flowViz
# load a flowSet to use as an example. library(flowFP)# load a flowSet to use as an example. library(flowFP)
These methods allow subset indexing for flowFP and flowFPPlex objects.
[ indexing returns a subset of the fingerprints in a flowFP.
[ indexing returns a subset of the fingerprints in a flowFP. If i matches one or more sampleClasses in the flowFP, the subset operator returns all fingerprints of that class (or classes). If no matches with sampleClass are found, then the operator attempts to match sampleNames, and returns all matches found.
[ indexing returns a new flowFPPlex with the specified subset of
fingerprints.
[[ indexing returns a single flowFP object from a flowplex.
This method joins one or more flowFP objects into a
flowFPPlex. Conceptually, each instance's fingerprint feature
vector is extended. An error will occur if the sample names and the class
names differ. If we imagine the fingerprints as a matrix where columns are
features and the rows are instances, this method binds the columns
onto the existing fingerprint matrix.
These methods append one or more fingerprints into a flowFPPlex.
append(x, values, after=length(x))
flowFPPlex methods:
Appends a fingerprint onto a flowFPPlex.
Appends a list of fingerprints to a flowFPPlex.
Combines two flowFPPlex-es into one.
Herb Holyst <[email protected]>, Wade Rogers <[email protected]>
flowFPPlex sampleNames sampleClasses
data(fs1) fp1 = flowFP(fs1, name="fingerprint 1") fp2 = flowFP(fs1, name="fingerprint 2") plex = flowFPPlex(fp1) plex = append(plex, fp2) plexdata(fs1) fp1 = flowFP(fs1, name="fingerprint 1") fp2 = flowFP(fs1, name="fingerprint 2") plex = flowFPPlex(fp1) plex = append(plex, fp2) plex
These methods return the bin boundaries as a list. The number of list elements
is equal to the number of features in the fingerprint.
Each boundary has two points, lower left (called ll) and upper right
(called ur). The dimensionality of ll and ur is the
dimensionality of the flowFrame or flowSet used to construct
the object.
binBoundary(object)
This method returns the bin boundaries from the underlying
flowFPModel.
This Method returns the flowFPModel bin boundaries.
These read-only methods retrieve data out of a fingerprint in three different transformations: “raw”, “normalized”, “log2norm”.
Transformations: There are three representations of the data that the user has access to, "raw" is the number of events in each bin, "normalized" is the ratio of the number of events divided by the expected number (the expected number of events is calculated by dividing the total number of events in the instance used to make the fingerprint by the number of features in the fingerprint). "log2norm" is the log2 of the normalized values.
counts(object, transformation=c("raw", "normalized", "log2norm")
This method returns a feature vector for each instance
in the flowFP, the user can change the way the features are represented,
by using the transformation argument. see below.
For a flowFPPlex the code assembles all of the
counts matrices from all of the flowFP contained in the flowFPPlex.
Constructor for flowFP-class.
flowFP(fcs, model = NULL, sampleClasses = NULL, ...)flowFP(fcs, model = NULL, sampleClasses = NULL, ...)
fcs |
A flowFrame or flowSet for which fingerprint(s) are desired. |
model |
A model generated with the flowFPModel constructor, or NULL. if NULL, a default model will be silently generated from all instances in x. |
sampleClasses |
An optional character vector describing modeling classes. If supplied, there must be exactly one element for each flowFrame in the flowSet in x (see Details). |
... |
If model is NULL, additional arguments are passed on to the model constructor. see flowFPModel for details. |
A flowFP object is a reduced representation of a flowFrame or flowSet with respect to a flowFPModel. The model is derived by finding multivariate regions containing (nearly) equal numbers of events in a training set using one or more user-specified parameters (e.g. CD45 and Side Scatter). The resulting flowFP represents the probability (i.e. number of events) of the instance(s) in x in each of the bins in the underlying model.
This representation is carried in a slot (called counts) in the object, a matrix in which each row corresponds ordinally to the flowFrames in x. Note that if x is a flowFrame and not a flowSet, the counts matrix will have only 1 row. In any case, each row of the matrix is an individual fingerprint, where the fingerprint elements (columns) have a one-to-one correspondence with the bins in the underlying model.
Another slot in the flowFP object describes the bin index for each event, allowing the user to select events according to their membership in bins. Think of this as "micro-gating".
An instance of an object of type flowFP.
Herb Holyst <[email protected]>, Wade Rogers <[email protected]>
M. Roederer, et. al. (2001) Probability Binning Comparison: A Metric for Quantitating Multivariate Distribution Differences, Cytometry 45, 47-55.
W. Rogers et. al. (2008) Cytometric Fingerprinting: Quantitative Characterization of Multivariate Distributions, Cytometry Part A 73, 430-441.
flowFP-class, flowFP, flowFPModel-class, flowFPModel
# load a flowSet to use as an example. # fs <- read.flowSet(your fcs files...) data(fs1) model <- flowFPModel(fs1, parameters=c(4,5), nRecursions=6) fp <- flowFP(fs1, model) plot(fp)# load a flowSet to use as an example. # fs <- read.flowSet(your fcs files...) data(fs1) model <- flowFPModel(fs1, parameters=c(4,5), nRecursions=6) fp <- flowFP(fs1, model) plot(fp)
This class represents fingerprints generated from a model derived from a flowFrame or flowSet.
Objects are created by calling the constructor flowFP.
counts:This matrix is organized such that rows correspond
to individual flowFrames, and columns to
the bins in the underlying flowFPModel. A
matrix element thus represents the number of events in
a single bin from a single flowFrame.
The sum of each row will equal the total number of events in
the corresponding flowFrame.
tags:This list specifies the bin number to which each event is assigned. There is a list element for each instance in the object. Each list element is a integer vector whose length is the number of events in the corresponding flowFrame.
sampleNames:Names of instances, taken from the flowSet on construction.
sampleClasses:Factor object describing the category of each instance. (e.g. "Normal" or "Diseased")
name:Inherited from flowFPModel.
parameters:Inherited from flowFPModel.
nRecursions:Inherited from flowFPModel.
trainingSet:Inherited from flowFPModel.
trainingSetParams:Inherited from flowFPModel.
dequantize:Inherited from flowFPModel.
split_val:Inherited from flowFPModel.
split_axis:Inherited from flowFPModel.
binBoundary:Inherited from flowFPModel.
.cRecursions:Inherited from flowFPModel.
.tmp_tags:Inherited from flowFPModel.
Class flowFPModel, directly.
[Used to select a subset of a fingerprint object.
countsReturns the counts matrix of a fingerprint object. See counts.
lengthReturns the number of instances in the fingerprint object.
nbinsReturns the number of bins (equal to 2^nRecursions) in the fingerprint object.
sampleClassesReturns the sampleClasses slot of the fingerprint object.
sampleClasses<-Set method for the sampleClasses slot of the fingerprint object. Note that the length of the provided object must be equal to the number of instances in the fingerprint object. If the provided object is not a factor it will be coerced to one.
tags Returns the tags slot. See tags.
showShows the contents of the fingerprint object.
summaryGives a summary of the fingerprint object.
Herb Holyst <[email protected]>, Wade Rogers <[email protected]>
M. Roederer, et. al. (2001) Probability Binning Comparison: A Metric for Quantitating Multivariate Distribution Differences, Cytometry 45, 47-55.
W. Rogers et. al. (2008) Cytometric Fingerprinting: Quantitative Characterization of Multivariate Distributions, Cytometry Part A 73, 430-441.
# load a flowSet to use as an example. library(flowFP) data(fs1) model <- flowFPModel(fs1, parameters=c(4,5), nRecursions=6) fp <- flowFP(fs1, model) plot(fp)# load a flowSet to use as an example. library(flowFP) data(fs1) model <- flowFPModel(fs1, parameters=c(4,5), nRecursions=6) fp <- flowFP(fs1, model) plot(fp)
A constructor for objects of type flowFPModel-class.
flowFPModel(fcs, name="Default Model", parameters=NULL, nRecursions="auto", dequantize=TRUE, sampleSize=NULL, excludeTime=TRUE)flowFPModel(fcs, name="Default Model", parameters=NULL, nRecursions="auto", dequantize=TRUE, sampleSize=NULL, excludeTime=TRUE)
fcs |
|
parameters |
A vector of parameters to be considered during model construction. |
nRecursions |
Number of times the FCS training data will be subdivided.
Each level generated |
dequantize |
If TRUE, all of the events in the training set will be made unique by adding a tiny value (proportional to the ordinal position of each event) to the data. |
sampleSize |
Used to specify the per-flowFrame sample size of the data to use in model generation. If NULL, all of the data in x is used. Setting this to a smaller number will speed up processing, at the cost of accuracy. |
name |
A descriptive name assigned to the model. |
excludeTime |
If TRUE (default) and no parameters are specified, the model constructor will atempt to exclude the time parameter from consideration when building the model. The time parameter is identified as have the label 'time', matched ignoring case. |
This function is used to create a flowFPModel, which can then be used to create a set of fingerprints using flowFP. A model is a representation of the multivariate probability density function for the training set x. This representation is in the form of a set of bins, each of which contains (nearly) the same number of events in the training set. Thus, the model can be regarded as a multivariate histogram of x with fixed bin count and variable bin size. Bins will be smallest in regions of high density, and largest in regions that are sparsely populated.
Bins are constructed by recursively subdividing the multivariate space specified
by parameters. At the first level, the entire space is divided in
half in the direction of the parameter with the highest variance. At the next
level, each of the halves from the first level is itself divided in half,
again in the direction of the parameter whose variance (among the events in the
parent bin) is the highest. Thus, for each level the number of resulting bins
is doubled. The number of bins in the model is thus .
A model is of little interest by itself. Its only utility is to provide a sort of basis function for forming fingerprints. It may be interesting in some cases to use the entire set x to build the model, and then to generate fingerprints for each instance in x. For this reason, calling flowFP without a model causes flowFPModel to be invoked silently under the hood in order to create a model of the flowSet provided to flowFP.
An object of type flowFPModel is returned.
If a flowSet is provided as input, it is internally collapsed to a flowFrame for binning.
Herb Holyst <[email protected]>, Wade Rogers <[email protected]>
M. Roederer, et. al. (2001) Probability Binning Comparison: A Metric for Quantitating Multivariate Distribution Differences, Cytometry 45, 47-55.
W. Rogers et. al. (2008) Cytometric Fingerprinting: Quantitative Characterization of Multivariate Distributions, Cytometry Part A 73, 430-441.
# load a flowSet to use as an example. library(flowFP) data(fs1) fs1 model <- flowFPModel(fs1, parameters=c(4,5), nRecursions=6) fp <- flowFP(fs1, model) plot(fp)# load a flowSet to use as an example. library(flowFP) data(fs1) fs1 model <- flowFPModel(fs1, parameters=c(4,5), nRecursions=6) fp <- flowFP(fs1, model) plot(fp)
This class is the fundamental clase for the package. It contains data and methods used to construct a model of the probability density function of a prototype dataset provided in the form of a flowFrame or flowSet.
Objects may only be created by calling the constructor function flowFPModel.
name:A user-supplied descriptive name for the model.
parameters:List of FCS parameters to use for model creation. Can be specified either by the names of the parameters or the indices of the parameters.
nRecursions:Number of levels of recursive subdivision.
The number of bins in the model will equal
.
trainingSet:Names of flowFrames from the FlowSet used to construct the model.
trainingSetParams:Names of all of the parameters from the flowFrames from the FlowSet used to construct the model.
dequantize:If TRUE, all of the event parameter values in the training set will be made unique by adding a tiny value (proportional to the ordinal position of each event) to the data.
split_val:A hairy array, aka list of vectors. Each list element is a vector representing the median values at which the data were split.
split_axis:A hairy array, aka list of vectors. Each list element is a vector representing the axis on which the data were split.
binBoundary:An object of class binBoundary, used to hold boundary information
used primarily for visualization.
.cRecursions:Private value to hold the number of levels of recursion used to
construct this model. Using nRecursions the resolution of a fingerprint
can be reduced, but it can never exceed this value.
.tmp_tags:Scratch array, total number of events in the training set long, that keeps track of the event's bin number. (this exists only to provide the underlying C function with a persistent scratch space it needs for bookkeeping. Not useful to the user.)
shows the contents of the model.
When creating a model you must keep in mind that it doesn't make sense to create more bins
(which is ) than the total number of events used to create the model.
The constructor checks for this.
When creating a model, you should specify only parameters that are common to all of the instances (flowFrames) in training data. For example, it does not make sense to compare PE from one flowFrame with FITC from another.
Herb Holyst <[email protected]>, Wade Rogers <[email protected]>
M. Roederer, et. al. (2001) Probability Binning Comparison: A Metric for Quantitating Multivariate Distribution Differences, Cytometry 45, 47-55.
W. Rogers et. al. (2008) Cytometric Fingerprinting: Quantitative Characterization of Multivariate Distributions, Cytometry Part A 73, 430-441.
flowFPModel - Constructor.
# load a flowSet to use as an example. library(flowFP) data(fs1) mod = flowFPModel(fs1)# load a flowSet to use as an example. library(flowFP) data(fs1) mod = flowFPModel(fs1)
This constructor creates a flowFPPlex, which is a collection of flowFPs.
flowFPPlex(fingerprints = NULL)flowFPPlex(fingerprints = NULL)
fingerprints |
List of flowFPs, or NULL. If NULL an empty flowFPPlex will be prepared, to which flowFPs may be added with append-methods. |
A flowFPPlex is a container object for a collection of flowFPs. It is useful in several ways. First, multiple-tube panels are often used to assay more reagents than can be done on a particular instrument. Second, it is sometimes useful to represent the same data with multiple flowFPModels.
Suppose that we have collected data from 1,000 patients, using an 8-tube panel. Imagine that Tubes 1 and 8 are isotype and viability tubes, respectively (we will ignore these tubes for now). The remaining Tubes 2-7 are of interest from a fingerprinting perspective. We wish to treat them as a unit. We might then create models that represent parameters in each of the tubes, across some (or all) of the patient samples (say, mod2, mod3, ..., mod7). We could then create corresponding flowFPs (say, fp2, fp3, ..., fp7). These can now be combined and treated as a single object of type flowFPPlex, as:
> plex <- flowFPPlex(c(fp2, fp3, fp4, fp5, fp6, fp7))
or if you prefer,> plex <- flowFPPlex (fingerprints=NULL)> plex <- append(plex, c(fp2, fp3, fp4, fp5, fp6, fp7))
The counts or density matrices can then be extracted simply using methods provided in flowFPPlex-class.
The second idea is to use multiple models to represent the same data. In this case we might create a model from, say, the "Normal" instances (call it mod\_norm), and another model from the "Cancer" instances (mod\_cancer). We might wish to do this to enhance the detection of regions of the distribution that are characteristically dominated by one type or the other. If our flowSet of all instances is called "fs1", then our two representations would be:
> fp_norm <- flowFP (fs1, mod_norm)> fp_cancer <- flowFP (fs2, mod_cancer)
and the plex is
plex <- flowFPPlex (c(fp_norm, fp_cancer))
Returns a flowFPPlex.
Herb Holyst <[email protected]>, Wade Rogers <[email protected]>
data(fs1) data(fs2) mod1 <- flowFPModel (fs1, parameters=c(2,5)) mod2 <- flowFPModel (fs2, parameters=c(2,5)) fp1_1 <- flowFP (fs1, mod1) fp1_2 <- flowFP (fs1, mod2) plex <- flowFPPlex(c(fp1_1, fp1_2))data(fs1) data(fs2) mod1 <- flowFPModel (fs1, parameters=c(2,5)) mod2 <- flowFPModel (fs2, parameters=c(2,5)) fp1_1 <- flowFP (fs1, mod1) fp1_2 <- flowFP (fs1, mod2) plex <- flowFPPlex(c(fp1_1, fp1_2))
This class holds a collection of flowFPs.
Objects should be created by calling the constructor function flowFPPlex.
fingerprints:Holds a list of flowFP objects.
Concatenates all of the counts for all of the fingerprints in the set.
returns the number of fingerprints in the set.
returns the number of flowFrames used to create this fingerprint.
returns the total number of features from the collection of fingerprints.
Herb Holyst <[email protected]>, Wade Rogers <[email protected]>
flowFPPlex - Constructor.
showClass("flowFPPlex")showClass("flowFPPlex")
An example FCS flowSet.
data(fs1)data(fs1)
This is an object of class flowSet composed of 7 flowFrames.
The flowSet corresponds to a Leukemia/Lymphoma panel for a sample. Each flowFrame correponds to one tube. The first one is an isotype control, and the rest are stained with different antibody cocktails. Parameter 5 (FL3 Log) is notable in that it is CD45-ECD for all of the tubes. This makes it possible to gate the entire collection of tubes using CD45 vs SSC. This set has a very similar distribution of CD45 vs SSC across all 7 tubes.
column names: FS Lin SS Log FL1 Log FL2 Log FL3 Log FL4 Log FL5 Log
Data were generously provided by Clarient, Inc.
W.T. Rogers and H. A. Holyst, “flowFP: A Bioconductor Package for Fingerprinting Flow Cytometric Data”, Submitted to: Advances in Bioinformatics, April 1, 2009.
data(fs1)data(fs1)
An example FCS flowSet.
data(fs2)data(fs2)
This is an object of class flowSet composed of 7 flowFrames.
The flowSet corresponds to a Leukemia/Lymphoma panel for a sample. Each flowFrame correponds to one tube. The first one is an isotype control, and the rest are stained with different antibody cocktails. Parameter 5 (FL3 Log) is notable in that it is CD45-ECD for all of the tubes. This makes it possible to gate the entire collection of tubes using CD45 vs SSC. In this set, the CD45 vs SSC distribution in Tube 5 differs significantly as compared with the other tubes.
column names: FS Lin SS Log FL1 Log FL2 Log FL3 Log FL4 Log FL5 Log
Data were generously provided by Clarient, Inc.
W.T. Rogers and H. A. Holyst, “flowFP: A Bioconductor Package for Fingerprinting Flow Cytometric Data”, Submitted to: Advances in Bioinformatics, April 1, 2009.
data(fs2)data(fs2)
This read-only accessor returns TRUE if the data object flowFP or flowFPPlex instances have been assigned a factor object representing the class membership of the instances.
hasClasses(object)
Returns TRUE if there are class assignments for each
instance in this flowFP.
If a flowFPPlex has classes associated with its
instances, they must all be the same.
This is a convenience function to check if the object is a flowFP.
is.flowFP(obj)is.flowFP(obj)
obj |
returns TRUE if the object is a |
returns TRUE if the object is a flowFP
data(fs1) fp = flowFP(fs1) is.flowFP(fp)data(fs1) fp = flowFP(fs1) is.flowFP(fp)
This is a convenience function to check if the object is a flowFPModel.
is.flowFPModel(obj)is.flowFPModel(obj)
obj |
returns TRUE if the object is a |
Returns TRUE is the object is a flowFPModel.
data(fs1) fp = flowFPModel(fs1) is.flowFPModel(fp)data(fs1) fp = flowFPModel(fs1) is.flowFPModel(fp)
This is a convenience function to check if the object is a flowFPPlex.
is.flowFPPlex(obj)is.flowFPPlex(obj)
obj |
returns TRUE if the object is a |
returns TRUE if the object is a flowFPPlex.
data(fs1) fp = flowFP(fs1) plex = flowFPPlex(fp) is.flowFPPlex(fp)data(fs1) fp = flowFP(fs1) plex = flowFPPlex(fp) is.flowFPPlex(fp)
This method sets or gets the number of fingerprints that have been concatinated together to form this flowFPPlex.
length(x)
Sets or gets the number of fingerprints in a flowFPPlex
The get method returns the number of flowFP objects in the plex. It is not the number of instances or the number of features in the counts matrix!
The set method can only truncate the set of flowFP objects from the right.
These methods return the model name(s) for an object.
name(object)
Returns the name of the model used to create these fingerprints.
Returns the model name.
Returns all of the model names found in the plex.
Theses methods return the number of leaf bins that were generated or will be generated by a specific model of plex of models.
nFeatures(object)
The number of bins in a fingerprint.
The number of bins that will be created by this model.
The total number of bins in a plex fingerprint.
Theses methods return the number of flowFrames that where used to create a fingerprint or plex.
nInstances(object)
The number of instances in a fingerprint.
The total number of instances in a plex fingerprint.
Gets or Sets the number of recursions for an object
nRecursions(object)
Sets and Gets the number of recursions for viewing a
flowFP object.
Gets the number of recursions for viewing a
flowFPModel object.
Gets the number of recursions for viewing a
flowFPPlex object.
When a model is computed (either with the flowFPModel constructor, or implicitly wit the flowFP constructor) the number of bins into which events are divided is determined. Because the binning process is recursive, given a high-resolution binning, any lower-resolution representation can easily be derived. The set methods do this derivation, and the get methods return the current resolution of the object.
Thus, the maximum value to which nRecursions may be set is determined by the value
of nRecursions at construction time. If an attempt is made to set nRecursions
higher than this value a warning will result.
These methods return the flow parameters that were taken into consideration at flowFPModel construction.
parameters(object)
Returns the parameters considered when creating fingerprint.
Returns the parameters that will be used to create a fingerprint.
An example of 96 well FCS flow data.
data(plate)data(plate)
This is an object of class flowSet comprising 96 flowFrames. Each flowFrame correponds to one sample. column names: FSC-H SSC-H FL1-H FL2-H FL3-H FL1-A FL4-H Time
This data set is derived from the references below. Each well in a 96-well plate is described by a flowFrame. Data were drastically down-sampled to 1000 events per flowFrame so that they could be included in this package. Note that flourescence parameters 5 (CD4 PerCP Cy5.5) and 7 (CD3 APC) are common across the entire set.
Original, non-sampled data are available at http://www.ficcs.org/software.html#Data_Files.
M. Inokuma, C. dela Rosa, C. Schmitt, P. Haaland, J. Siebert, D. Petry, M. Tang, M. A. Suni, S. A. Ghanekar, D. Gladding, J. F. Dunne, V. C. Maino, M. L. Disis, and H. T. Maecker. “Functional t cell responses to tumor antigens in breast cancer patients have a distinct phenotype and cytokine signature”. J Immunol, 179(4):2627-33, Aug 15 2007.
M. Inokuma, C. dela Rosa, C. Schmitt, P. Haaland, J. Siebert, D. Petry, M. Tang, M. A. Suni, S. A. Ghanekar, D. Gladding, J. F. Dunne, V. C. Maino, M. L. Disis, and H. T. Maecker, 2008. Data available at http://www.ficcs.org/software.html#Data_Files.
data(plate)data(plate)
These methods allow the user to plot flowFP objects with a number of options.
plot (x, y, ...)
x = "flowFPModel", y = "missing" Visualize a flowFPModel.
Optional Args: (parameters=NULL, alpha=1, border="gray",
showbins=1:nFeatures(x), ylim=NULL, xlim=NULL, main="Model", ...)
x = "flowFPModel", y = "flowFrame" Visualize a flowFPModel along with a flowFrame.
Optional Args: (parameters=NULL, alpha=1, border="gray",
showbins=1:nFeatures(x), ylim=NULL, xlim=NULL, main="Model", ...)
x = "flowFPModel", y = "flowSet" Visualize a flowFPModel along with a flowSet.
Optional Args: (parameters=NULL, alpha=1, border="gray",
showbins=1:nFeatures(x), ylim=NULL, xlim=NULL, main="Model", ...)
x = "flowFP", y = "missing" Visualize Fingerprints.
Optional Args: (type=c("tangle", "stack", "grid", "qc", "plate"), ...)
See Notes for additional details.
x = "flowFP", y = "flowFrame"Visualize a single fingerprint with a flowFrame. See Notes.
x = "flowFP", y = "flowSet" Visualize Fingerprints with a flowSet.
Optional Args: (x, y, showbins=NULL, showfp=TRUE, ...)
x = "flowFPPlex", y = "missing" Visualize Fingerprints in a flowFPPlex.
Optional Args: (type=c("tangle", "stack", "grid","qc", "plate"), ...)
parametersFlow parameters involved in fingerprinting.
showbinsWhich bins to render. May be an integer, list of integers, or boolean. Default: 1:nFeatures(x).
showfpBoolean. Do we want to see the fingerprints (TRUE) or just the parameter plot (FALSE).
xlimRange of values for the x-axis. Default: range of the data.
ylimRange of values for the y-axis. Default: range of the data.
typeSpecify the type of plot. Current values supported are: “tangle”, “stack”, “grid”, “qc”, “plate”.
transformationSpecify transformation. Values may be: “raw”, “normalized”, “log2norm”
useClassesBoolean. Colors lines by sampleClass. Default=FALSE.
linecolsSpecify color(s) for fingerprint lines. Default = ‘black’. When
used in conjunction with useClasses, this can be a list of colors whose length matches
the number of levels in the sampleClasses factor object.
alphaOpacity of rectangles representing bins.
borderColor to draw the border of a bin. Default: ‘gray’.
highlightWhich fingerprints to highlight. Default=1:length(x).
mainTitle for the plot. Default: various.
xlabLabel for the x-axis.
vert_scaleSets the vertical scale for some plots.
respectBoolean. If TRUE, respect a square aspect ratio.
methodDetermine QC metric for fingerprint deviation. Either sd (standard deviation)
or max (maximum deviation).
red_limitValue corresponding to a red color for the QC metric.
...Other parameters passed through to generic plot.
generic plot args such as pch and cex can be
used to visualize events in specific bins.
Optional Args: (transformation=c("raw", "normalized", "log2norm"),
linecols=NULL, highlight=NULL, ylim=NULL,
useClasses=FALSE, main="Fingerprints", xlab='Feature Index', ...)
Optional Args: (transformation=c("raw", "normalized", "log2norm"),
linecols=NULL, useClasses=FALSE, vert_scale=3,
ylim=NULL, main="Fingerprints",...)
Optional Args: (vert_scale=3, main="Fingerprints", linecols="black",
transformation=c("raw", "normalized", "log2norm"),
respect=FALSE)
Optional Args: (main="Fingerprint Deviation Plot",
transformation=c("log2norm", "raw", "normalized"),
vert_scale=3, method=c("sd", "max"), red_limit=1.0,
respect=FALSE)
Optional Args: (main="Fingerprint Deviation Plot",
transformation=c("log2norm", "raw", "normalized"),
vert_scale=3, method=c("sd", "max"), red_limit=1.0)
These methods get or set the sample classes for a fingerprint or a set of fingerprints.
sampleClasses(object)
Returns the sampleClass factor object found in a flowFP
Returns the sampleClass factor object found in a flowFPPlex
These methods get the sample names for a fingerprint or a set of fingerprints.
sampleNames(object)
returns the sample names in a flowFP object.
returns the sample names in a flowFPPlex object.
These methods give basic information about the contents of a flowFP object.
show(object)
Show the sample names and model information for a flowFP.
Shows the samples used in creating the model.
Shows all of the sample named and all of the models found in the plex.
These methods give a summary of the flowFP object.
summary(object)
Show the per-instance min/max/mean counts in aflowFP.
Shows length, class and mode info for a flowFPModel.
Shows length, class and mode info for a flowFPPlex.
The tags method is a map into a flowFP object which defines the bin number
to which each event in the object has been assigned in fingerprinting process.
tags(object)
returns a list of vectors.
This method returns a list of vectors. There is a one-to-one correspondence between
vectors in the list and flowFrames in the flowSet in the flowFP.
For each vector in the list, there is a one-to-one correspondence between the vector elements and the events in the corresponding flowFrame. Each vector element is an integer index that indicates the bin number to which that event was assigned in the fingerprinting process.