Package 'Mulcom' reference manual

Title:	Calculates Mulcom test
Description:	Identification of differentially expressed genes and false discovery rate (FDR) calculation by Multiple Comparison test.
Authors:	Claudio Isella
Maintainer:	Claudio Isella <[email protected]>
License:	GPL-2
Version:	1.57.0
Built:	2024-10-30 08:23:53 UTC
Source:	https://github.com/bioc/Mulcom

Affy Dataset

Description

Affy Dataset

Author(s)

Claudio Isella, [email protected]

cross mapping table

Description

cross mapping table

Author(s)

Claudio Isella, [email protected]

MulCom Harmonic Mean

Description

Computes harmonic means across groups replicate Should not be called directly

Usage

harmonicMean(index)
harmonicMean(index)

Arguments

index

a numeric vector with the groups labels of the samples. 0 are the control samples. Number must be progressive

Details

harmonicMean calculates harmonic means across groups replicate for the estimation of Mulcom Test

Value

a numeric vector

Author(s)

Claudio Isella, [email protected]

References

[email protected]

Illumina Dataset

Description

Illumina Dataset

Author(s)

Claudio Isella, [email protected]

Ilmn Dataset

Description

Ilmn Dataset

Author(s)

Claudio Isella, [email protected]

significant gene list with limma in Affymetrix

Description

significant gene list with limma in Affymetrix

Author(s)

Claudio Isella, [email protected]

significant gene list with limma in Illumina

Description

significant gene list with limma in Illumina

Author(s)

Claudio Isella, [email protected]

MulCom Calculation

Description

Calculates MulCom test score for given m and t parameters

Usage

mulCalc(Mulcom_P, m, t)
mulCalc(Mulcom_P, m, t)

Arguments

`Mulcom_P`	an object of class MULCOM
`m`	m: a numeric value corresponding to log 2 ratio correction for MulCom Test
`t`	t: a numeric value corresponding to T values for MulCom Test

Details

mulCalc Calculate the Mulcom Score with m and t defined by the user

Mulcom_P:: an object of class MULCOM_P

m:: a number corresponding to log 2 ratio correction for MulCom Test

t:: a number corresponding to T values for MulCom Test

Author(s)

Claudio Isella, [email protected]

Examples

data(benchVign)
mulcom_scores <- mulScores(Affy, Affy$Groups)
mulcom_calc <- mulCalc(mulcom_scores, 0.2, 2)
data(benchVign)
mulcom_scores <- mulScores(Affy, Affy$Groups)
mulcom_calc <- mulCalc(mulcom_scores, 0.2, 2)

Identify the Mulcom candidate feature selection

Description

Identify the Mulcom candidate feature selection by the m and T defined by the user

Usage

mulCAND(eset, Mulcom_P, m, t, ese = "T")
mulCAND(eset, Mulcom_P, m, t, ese = "T")

Arguments

`eset`	an `AffyBatch`
`Mulcom_P`	an object of class MULCOM
`m`	m: a numeric vector corresponding to log 2 ratio correction
`t`	t: a numeric vector corresponding to the MulCom T values
`ese`	True or False

Details

mulCAND Identify the Mulcom candidate feature selection by the m and T defined by the user

Author(s)

Claudio Isella, [email protected]

Examples

data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10,2)
mulcom_cand <- mulCAND(Affy, mulcom_perm, 0.2, 2)
data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10,2)
mulcom_cand <- mulCAND(Affy, mulcom_perm, 0.2, 2)

Class MulCom Permutation

Description

This is a class representation MulCom test scores permutation

Objects from the Class

Objects can be created using the function mulScores on ExpressionSet.

Slots

FC:: Object of class numeric representing delta between all experimental groups and the reference groups
MSE_Corrected:: Object of class numeric representing the MulCom test estimation of mean square error as described in the formula of the Dunnett's t-test
FCp:: Object of class numeric representing delta between all experimental groups and the reference groups in permutated data
MSE_Correctedp:: Object of class numeric representing the MulCom test estimation of mean square error as described in the formula of the Dunnett's t-test in permutated data

Author(s)

Claudio Isella

Examples

data(benchVign)
mulcom_scores <- mulScores(Affy, Affy$Groups)
data(benchVign)
mulcom_scores <- mulScores(Affy, Affy$Groups)

Class MulCom

Description

This is a class representation MulCom test scores

Objects from the Class

Objects can be created using the function mulScores on ExpressionSet.

Slots

FC:: Object of class numeric representing difference between all experimental groups and the reference groups
HM:: Object of class numeric representing the harmonic means in all subgroups
MSE_Corrected:: Object of class numeric representing the MulCom test estimation of mean square error as described in the formula of the Dunnett's t-test

Author(s)

Claudio Isella

Examples

data(benchVign)
mulcom_scores <- mulScores(Affy, Affy$Groups)
data(benchVign)
mulcom_scores <- mulScores(Affy, Affy$Groups)

significant gene list with limma in Illumina

Description

significant gene list with limma in Illumina

Author(s)

Claudio Isella, [email protected]

MulCom Delta

Description

Computes Delta for all the experimental points in the datasets in respect to control Should not be called directly

Usage

mulDELTA(vector, index)
mulDELTA(vector, index)

Arguments

`vector`	vector: numeric vector with data measurements
`index`	a numeric vector with the labels of the samples. 0 are the control samples. number must be progressive

Details

mulDELTA An internal function that should not be called directly. It calculates differential expression in the groups defined in the index class vector, in respect to the 0 groups

Value

`vector`	a numeric vector with data measurements
`index`	a numeric vector with the labels of the samples. 0 are the control samples. number must be progressive

Author(s)

Claudio Isella, [email protected]

Examples

data(benchVign)
mulcom_delta <- mulDELTA(exprs(Affy[1,]),Affy$Groups)

data(benchVign)
mulcom_delta <- mulDELTA(exprs(Affy[1,]),Affy$Groups)

MulCom Test Differential analysis

Description

Identify the differentially expressed features for a specific comparison with given m and t value

Usage

mulDiff(eset, Mulcom_P, m, t, ind)
mulDiff(eset, Mulcom_P, m, t, ind)

Arguments

`eset`	An ExpressionSet object from package Biobase
`Mulcom_P`	An object of class Mulcom_P
`m`	the m values for the analysis
`t`	the t values for the analysis
`ind`	and index refeing to te comparison, should be numeric

Value

`eset`	An ExpressionSet object from package Biobase
`Mulcom_P`	An object of class Mulcom_P
`m`	the m values for the analysis
`t`	the t values for the analysis
`ind`	and index refeing to te comparison, should be numeric

Author(s)

Claudio Isella, [email protected]

Examples

data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10, 7)
mulcom_diff <- mulDiff(Affy, mulcom_perm, 0.2, 2)
data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10, 7)
mulcom_diff <- mulDiff(Affy, mulcom_perm, 0.2, 2)

MulCom False Significant Genes

Description

Calculate the False Significant Genes for m and t defined by the user

Usage

mulFSG(Mulcom_P, m, t)
mulFSG(Mulcom_P, m, t)

Arguments

`Mulcom_P`	an object of class MULCOM
`m`	m: a numeric value corresponding to log 2 ratio correction for MulCom Test
`t`	t: a numeric value corresponding to t values for MulCom Test

Details

mulFDR evaluate the False Significant genes on the Mulcom_P object according to specific m and t parameters. For each permutation it is calculated the number of positive genes. An estimation of the false called genes is evaluated with the median for each experimental subgroups

Author(s)

Claudio Isella, [email protected]

Examples

data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10, 7)
mulcom_fsg <- mulFSG(mulcom_perm, 0.2, 2)
data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10, 7)
mulcom_fsg <- mulFSG(mulcom_perm, 0.2, 2)

Mulcom Index for Monte Carlo Simlation

Description

Random assebly of the groups indices for Monte Carlo Simulation

Usage

mulIndex(index, np, seed)
mulIndex(index, np, seed)

Arguments

`index`	the vector with the groups of analysis, must be numeric and 0 correspond to the reference.
`np`	number of permutation in the simulation
`seed`	seed for permtations

Details

'mulIndex' generates random index for the function mulPerm. it is not directly called by the user.

Value

A matrix with all indices permutations

Author(s)

Claudio Isella, [email protected]

Examples

data(benchVign)
mulcom_scores <- mulIndex(Affy$Groups, 5, 7)
data(benchVign)
mulcom_scores <- mulIndex(Affy$Groups, 5, 7)

generates a consensus matrix from list of genes

Description

generates a consensus matrix from list of genes

Usage

mulInt(...)
mulInt(...)

Arguments

...

the function requires vector files as imputs

Details

mulCAND generates a consensus matrix from list of genes

Author(s)

Claudio Isella, [email protected]

Examples

data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10,2)
mulcom_opt <- mulOpt(mulcom_perm, vm = seq(0,0.5, 0.1), vt = seq(1,3, 0.1))

h1_opt <- mulParOpt(mulcom_perm, mulcom_opt, ind = 1, th = 0.05)
h2_opt <- mulParOpt(mulcom_perm, mulcom_opt, ind = 1, th = 0.05)

int <- mulInt(h1_opt, h2_opt)

data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10,2)
mulcom_opt <- mulOpt(mulcom_perm, vm = seq(0,0.5, 0.1), vt = seq(1,3, 0.1))

h1_opt <- mulParOpt(mulcom_perm, mulcom_opt, ind = 1, th = 0.05)
h2_opt <- mulParOpt(mulcom_perm, mulcom_opt, ind = 1, th = 0.05)

int <- mulInt(h1_opt, h2_opt)

MulCom Mean Square Error

Description

Computes Mean Square Error for all the experimental points in the datasets in respect to control. should not be called directly

Usage

mulMSE(vector, index, tmp = vector())
mulMSE(vector, index, tmp = vector())

Arguments

`vector`	a numeric vector with data mesurements
`index`	a numeric vector with the labels of the samples. 0 are the control samples. number must be progressive
`tmp`	a vector

Details

mulMSE An internal function that should not be called directly. It calculates within group means square error for the values defined in the x vector according to the index class vector

Value

`vector`	a numeric vector with data measurements
`index`	a numeric vector with the labels of the samples. 0 are the control samples. number must be progressive
`tmp`	a vector

Author(s)

Claudio Isella, [email protected]

MulCom optimization

Description

The function systematically performs the calculation of significant genes and corresponding FDR for all the combination of given list of m and t values.

Usage

mulOpt(Mulcom_P, vm, vt) 
mulOpt(Mulcom_P, vm, vt)

Arguments

`Mulcom_P`	an object of class Mulcom_P
`vm`	a vector of m values to test
`vt`	a vector of t values to test

Details

mulOpt The function systematically performs the calculation of significant genes and corresponding FDR for all the combination of given list of m and t values.

Author(s)

Claudio Isella, [email protected]

Examples

data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10, 7)
mulcom_opt <- mulOpt(mulcom_perm, seq(0.1, 0.5, 0.1), seq(1, 3, 0.1))
data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10, 7)
mulcom_opt <- mulOpt(mulcom_perm, seq(0.1, 0.5, 0.1), seq(1, 3, 0.1))

MulCom Parameter Optimization

Description

Function to optimize Mulcom parameter for maximim nuber of genes with a user defined FDR

Usage

   mulOptPars(opt, ind, ths) 
mulOptPars(opt, ind, ths)

Arguments

`opt`	an MulCom optimization object
`ind`	index corresponding to the comparison
`ths`	a threshold for the FDR optimization, default is 0.05

Details

mulOptPars MulCom optimization function to identify best parameters

Author(s)

Claudio Isella, [email protected]

Examples

data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10, 7)
#mulcom_opt <- mulOpt(mulcom_perm, seq(0.1, 0.5, 0.1), seq(1, 3, 0.1))
#optThs <- mulOptPars(mulcom_opt, 1, 0.05)
data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10, 7)
#mulcom_opt <- mulOpt(mulcom_perm, seq(0.1, 0.5, 0.1), seq(1, 3, 0.1))
#optThs <- mulOptPars(mulcom_opt, 1, 0.05)

MulCom optimization Plot

Description

MulCom optimization Plot to identify best configuration paramters

Usage

   mulOptPlot(M.Opt, ind, th, smooth = "NO") 
mulOptPlot(M.Opt, ind, th, smooth = "NO")

Arguments

`M.Opt`	an MulCom optimization object
`ind`	index corresponding to the comparison to plot
`th`	a threshold for the FDR plot
`smooth`	indicates whether the FDR plot will show a significant threshold or will be continuous.

Details

mulOptPlot MulCom optimization Plot

Value

a numeric vector

Author(s)

Claudio Isella, [email protected]

Examples

   data(benchVign)
   mulcom_perm <- mulPerm(Affy, Affy$Groups, 10,2)
   mulcom_opt <- mulOpt(mulcom_perm, vm=seq(0.1, 0.5, 0.1), vt=seq(1, 3,1))
   mulOptPlot(mulcom_opt, 1, 0.05)
data(benchVign)
   mulcom_perm <- mulPerm(Affy, Affy$Groups, 10,2)
   mulcom_opt <- mulOpt(mulcom_perm, vm=seq(0.1, 0.5, 0.1), vt=seq(1, 3,1))
   mulOptPlot(mulcom_opt, 1, 0.05)

MulCom Parameters Optimization

Description

MulCom parameter optimization function to identify best combination of t and m providing maximum number of genes at a given FDR

Usage

   mulParOpt(perm, M.Opt, ind, th, image = "T")
mulParOpt(perm, M.Opt, ind, th, image = "T")

Arguments

`perm`	a object with permutated MulCom Scores
`M.Opt`	an MulCom optimization object
`ind`	index corresponding to the comparison to plot
`th`	a threshold for the FDR plot
`image`	default = "T", indicates is print the MulCom optimization plot

Details

mulParOpt The function mulParOpt is designed to identify the optimal m and t values combination leading to the maximum number of differentially regulated genes satisfying an user define FDR threshold. In case of equal number of genes, the combination of m and t with the lower FDR will be prioritized. In case of both identical number of genes and FDR, the function will chose the highest t. The function optionally will define a graphical output to visually inspect the performance of the test at given m and t parameters for a certain comparison.

Author(s)

Claudio Isella, [email protected]

Examples

   data(benchVign)
   mulcom_perm <- mulPerm(Affy, Affy$Groups, 10,2)
   mulcom_opt <- mulOpt(mulcom_perm, vm=seq(0.1, 0.5, 0.1), vt=seq(1, 3,1))
   mulParOpt(mulcom_perm, mulcom_opt, 1, 0.05)
data(benchVign)
   mulcom_perm <- mulPerm(Affy, Affy$Groups, 10,2)
   mulcom_opt <- mulOpt(mulcom_perm, vm=seq(0.1, 0.5, 0.1), vt=seq(1, 3,1))
   mulParOpt(mulcom_perm, mulcom_opt, 1, 0.05)

MulCom Permutation

Description

Reiterate MulCom Test on permutated data to perform Montecarlo simulation

Usage

mulPerm(eset, index, np, seed, segm = "F")
mulPerm(eset, index, np, seed, segm = "F")

Arguments

`eset`	An an `AffyBatch` object, each row of must correspond to a variable and each column to a sample.
`index`	a numeric vector of length ncol(data) with the labels of the samples. 0 are the reference samples.
`np`	a numeric values indicating the number of permutation to perform. It is set as default to 10
`seed`	set the seed of the permutaton, default is 1
`segm`	a default set to F. This parametheres requires to be setted to avoid segmentation fault of C subroutin in the case of very large datasets.

Details

mulPerm

Author(s)

Claudio Isella, [email protected]

Examples

data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10,2)
data(benchVign)
mulcom_perm <- mulPerm(Affy, Affy$Groups, 10,2)

MulCom Permutation

Description

R pipe to C function not called directly by user that reiterate MulCom Test on permutated data to perform Monte Carlo simulation

Usage

mulPermC(eset, index, means, mse, n, m, nump, ngroups, reference)
mulPermC(eset, index, means, mse, n, m, nump, ngroups, reference)

Arguments

`eset`	An an `AffyBatch` object, each row of must correspond to a variable and each column to a sample.
`index`	a numeric vector of length ncol(data) with the labels of the samples. 0 are the reference samples.
`means`	entry for the means output.
`mse`	entry for the mean square errors output
`n`	number of rows in obext of class eset
`m`	number of columns
`nump`	number of permutation to perform
`ngroups`	a number corresponding to the number of groups in the analysis.
`reference`	reference for the comparisons. typically it is 0

Details

mulPerm

Author(s)

Claudio Isella, [email protected]

Examples

data(benchVign)
data(benchVign)

MulCom Score Calculation

Description

Computes the scores for the MulCom test. The function calculates the numerator and the denominator of the test without the parameters m and t

Usage

mulScores(eset, index)
mulScores(eset, index)

Arguments

`eset`	An an `AffyBatch` object, each row of must correspond to a variable and each column to a sample.
`index`	a numeric vector of length ncol(data) with the labels of the samples. 0 are the reference samples.

Details

'mulScore' computes the scores for the MulCom test for multiple point profile. The Mulcom test is designed to compare each experimental mean with the control mean and it is derived from the "Dunnett's test". Dunnett's test controls the Experiment-wise Error Rate and is more powerful than tests designed to compare each mean with each other mean. The test is conducted by computing a modified t-test between each experimental group and the control group.

Value

An Object of class MULCOM from Mulcom package

Author(s)

Claudio Isella, [email protected]

Examples

data(benchVign)
mulcom_scores <- mulScores(Affy, Affy$Groups)
data(benchVign)
mulcom_scores <- mulScores(Affy, Affy$Groups)

MulCom Sum of Square Error

Description

Computes sum of square errors for all the experimental points in the datasets Should not be called directly

Usage

mulSSE(vec, index)
mulSSE(vec, index)

Arguments

`vec`	a numeric vector with data measurements
`index`	a numeric vector with the labels of the samples. 0 are the control samples. number should be progressive

Details

mulSSE An internal function that should not be called directly. It calculates sum of square error in the groups defined in the index class vector.

Value

`vec`	a numeric vector with data measurements
`index`	a numeric vector with the labels of the samples. 0 are the control samples. number must be progressive

Author(s)

Claudio Isella, [email protected]

significant gene list with SAM in Affymetrix

Description

significant gene list with SAM in Affymetrix

Author(s)

Claudio Isella, [email protected]

significant gene list with SAM in Illumina

Description

significant gene list with SAM in Illumina

Author(s)

Claudio Isella, [email protected]

sam Parameter Optimization

Description

Function to optimize Sam parameter for maximim nuber of genes with a user defined FDR

Usage

samOptPars(opt, ths) 
samOptPars(opt, ths)

Arguments

`opt`	an Sam optimization object
`ths`	a threshold for the FDR optimization

Value

a numeric vector

Author(s)

Claudio Isella, [email protected]

Package 'Mulcom'

Help Index

Affy Dataset

Description

Author(s)

cross mapping table

Description

Author(s)

MulCom Harmonic Mean

Description

Usage

Arguments

Details

Value

Author(s)

References

Illumina Dataset

Description

Author(s)

Ilmn Dataset

Description

Author(s)

significant gene list with limma in Affymetrix

Description

Author(s)

significant gene list with limma in Illumina

Description

Author(s)

MulCom Calculation

Description

Usage

Arguments

Details

Author(s)

Examples

Identify the Mulcom candidate feature selection

Description

Usage

Arguments

Details

Author(s)

Examples

Class MulCom Permutation

Description

Objects from the Class

Slots

Author(s)

Examples

Class MulCom

Description

Objects from the Class

Slots

Author(s)

Examples

significant gene list with limma in Illumina

Description

Author(s)

MulCom Delta

Description

Usage

Arguments

Details

Value

Author(s)

Examples

MulCom Test Differential analysis

Description

Usage

Arguments

Value

Author(s)

Examples

MulCom False Significant Genes

Description

Usage

Arguments

Details

Author(s)

Examples

Mulcom Index for Monte Carlo Simlation