Engine-class {Umpire}

The "Engine" Class

Description

The Engine class is a tool (i.e., an algorithm) used to simulate vectors of gene expression from some underlying distribution.

Usage

Engine(components)
## S4 method for signature 'Engine'
nComponents(object, ...)
## S4 method for signature 'Engine'
alterMean(object, TRANSFORM, ...)
## S4 method for signature 'Engine'
alterSD(object, TRANSFORM, ...)
## S4 method for signature 'Engine'
nrow(x)
## S4 method for signature 'Engine'
rand(object, n, ...)
## S4 method for signature 'Engine'
summary(object, ...)

Arguments

components

object of class list, each element of which contains the parameters for the underlying distribution that the gene expression follows. A component can be viewed as a special case of an engine that only has one component.

object, x

object of class Engine

TRANSFORM

function takes as its input a vector of mean expression or standard deviation and returns a transformed vector that can be used to alter the appropriate slot of the object.

n

numeric scalar representing number of samples to be simulated

...

extra arguments for generic or plotting routines

Details

In most cases, an engine object is an instantiation of a more general family or class that we call an ABSTRACT ENGINE. Every abstract engine is an ordered list of components, which can also be thought of as an engine with parameters. We instantiate an engine by binding all the free parameters of an abstract engine to actual values. Note that partial binding (of a subset of the free parameters) produces another abstract engine.

For all practical purposes, a COMPONENT should be viewed as an irreducible abstract engine. Every component must have an IDENTIFIER that is unique within the context of its enclosing abstract engine. The identifer may be implicitly taken to be the position of the component in the ordered list.

We interpret an Engine as the gene expression generator for a homogenous population; effects of cancer on gene expression are modeled at a higher level.

Value

The Engine generator returns an object of class Engine.

The alterMean method returns an object of class Engine with altered mean.

The alterSD method returns an object of class Engine with altered standard deviation.

The nrow method returns the number of genes (i.e, the length of the vector) the Engine object will generate.

The rand method returns nrow(Engine)*n matrix representing the expressions of nrow(Engine) genes and n samples.

The summary method prints out the number of components included in the Engine object.

The nComponents method returns the number of components in the Engine object.

Objects from the Class

Objects can be created by calls of the form new("Engine", components=components), or use the Engine generator function. Every engine is an ordered list of components, which generates a contiguous subvector of the total vector of gene expression.

Methods

alterMean(object, TRANSFORM, ...)

Takes an object of class Engine, loops over the components in the Engine, alters the mean as defined by TRANSFORM function, and returns a modified object of class Engine.

alterSD(object, TRANSFORM, ...)

Takes an object of class Engine, loops over the components in the Engine, alters the standard deviation as defined by TRANSFORM function, and returns a modified object of class Engine.

nrow(x)

Counts the total number of genes (i.e, the length of the vector the Engine will generate).

rand(object, n, ...)

Generates nrow(Engine)*n matrix representing gene expressions of n samples following the underlying distribution captured in the object of Engine.

summary(object, ...)

Prints out the number of components included in the object of Engine.

Author(s)

Kevin R. Coombes [e-mail], Jiexin Zhang [e-mail],

References

Zhang J, Coombes KR.
Sources of variation in false discovery rate estimation include sample size, correlation, and inherent differences between groups.
BMC Bioinformatics. 2012; 13 Suppl 13:S1.

See Also

EngineWithActivity

Examples

showClass("Engine")
nComp <- 10
nGenes <- 100
comp <- list()
for (i in 1:nComp) {
  comp[[i]] <- IndependentNormal(rnorm(nGenes/nComp, 6, 1.5),
                                 1/rgamma(nGenes/nComp, 44, 28))
}
myEngine <- Engine(comp)
nrow(myEngine)
nComponents(myEngine)
summary(myEngine)
myData <- rand(myEngine, 5)
dim(myData)
summary(myData)
OFFSET <- 2
myEngine.alterMean <- alterMean(myEngine, function(x){x+OFFSET})
myData.alterMean <- rand(myEngine.alterMean, 5)
summary(myData.alterMean)
SCALE <- 2
myEngine.alterSD <- alterSD(myEngine, function(x){x*SCALE})
myData.alterSD <- rand(myEngine.alterSD, 5)
summary(myData.alterSD)