Package 'highMLR' reference manual

Title:	Feature Selection for High Dimensional Survival Data
Description:	Perform high dimensional Feature Selection in the presence of survival outcome. Based on Feature Selection method and different survival analysis, it will obtain the best markers with optimal threshold levels according to their effect on disease progression and produce the most consistent level according to those threshold values. The functions' methodology is based on by Sonabend et al (2021) <doi:10.1093/bioinformatics/btab039> and Bhattacharjee et al (2021) <arXiv:2012.02102>.
Authors:	Atanu Bhattacharjee [aut, cre, ctb], Gajendra K. Vishwakarma [aut, ctb], Souvik Banerjee [aut, ctb]
Maintainer:	Atanu Bhattacharjee <[email protected]>
License:	GPL-3
Version:	0.1.1
Built:	2024-11-09 04:52:58 UTC
Source:	https://github.com/cran/highMLR

High dimensional head and neck cancer survival and gene expression data

Description

High dimensional head and neck cancer gene expression data

Usage

hnscc
hnscc

Format

A dataframe with 565 rows and 104 variables

ID: "Column/Variable name" consisting id of subjects
Death: "Column/Variable name" consisting survival event
OS: "Column/Variable name" consisting duration of overall survival
PFS: "Column/Variable name" consisting duration of progression free survival
Prog: "Column/Variable name" consisting progression event
GJB1,...,HMGCS2: High dimensional covariates

Examples

data(hnscc)
data(hnscc)

Applications of machine learning in survival analysis by prognostic classification of genes by CoxPH model.

Description

Applications of machine learning in survival analysis by prognostic classification of genes by CoxPH model.

Usage

mlclassCox(m, n, idSurv, idEvent, Time, s_ID, per = 20, fold = 3, data)
mlclassCox(m, n, idSurv, idEvent, Time, s_ID, per = 20, fold = 3, data)

Arguments

`m`	Starting column number from where high dimensional variates to be selected.
`n`	Ending column number till where high dimensional variates to be selected.
`idSurv`	"Column/Variable name" consisting duration of survival.
`idEvent`	"Column/Variable name" consisting survival event.
`Time`	"Column/Variable name" consisting Times of repeated observations.
`s_ID`	"Column/Variable name" consisting unique identification for each subject.
`per`	Percentage value for ordering, default=20.
`fold`	Number of folds for re-sampling, default=3.
`data`	High dimensional data containing survival observations with multiple covariates.

Value

A list of genes as per their classifications

GeneClassification: List of genes classified using Cox proportional hazard model
GeneClassification$Positive_Gene: Sublist of genes classified as positive genes
GeneClassification$Negative_Gene: Sublist of genes classified as negative genes
GeneClassification$Volatile_Gene: Sublist of genes classified as volatile genes
Result: A dataframe consisting threshold values with corresponding coefficients and p-values.

Examples

## Not run: 
data(srdata)
mlclassCox(m=50,n=59,idSurv="OS",idEvent="event",Time="Visit",s_ID="ID",per=20,fold=3,data=srdata)

## End(Not run)
## Not run: 
data(srdata)
mlclassCox(m=50,n=59,idSurv="OS",idEvent="event",Time="Visit",s_ID="ID",per=20,fold=3,data=srdata)

## End(Not run)

Applications of machine learning in survival analysis by prognostic classification of genes by Kaplan-Meier estimator.

Description

Applications of machine learning in survival analysis by prognostic classification of genes by Kaplan-Meier estimator.

Usage

mlclassKap(m, n, idSurv, idEvent, Time, s_ID, per = 20, fold = 3, data)
mlclassKap(m, n, idSurv, idEvent, Time, s_ID, per = 20, fold = 3, data)

Arguments

`m`	Starting column number from where high dimensional variates to be selected.
`n`	Ending column number till where high dimensional variates to be selected.
`idSurv`	"Column/Variable name" consisting duration of survival.
`idEvent`	"Column/Variable name" consisting survival event.
`Time`	"Column/Variable name" consisting timepoints of repeated observations.
`s_ID`	"Column/Variable name" consisting unique identification for each subject.
`per`	Percentage value for ordering, default=20.
`fold`	Number of fold for resampling, default=3.
`data`	High dimensional data containing survival observations and high dimensional covariates.

Value

A list of genes as per their classifications

GeneClassification: List of genes classified using Cox proportional hazard model
GeneClassification$Positive_Gene: Sublist of genes classified as positive genes
GeneClassification$Negative_Gene: Sublist of genes classified as negative genes
GeneClassification$Volatile_Gene: Sublist of genes classified as volatile genes
Result: A dataframe consisting threshold values with corresponding coefficients and p-values.

Examples

 ## Not run: 
##
mlclassKap(m=50,n=59,idSurv="OS",idEvent="event",Time="Visit",s_ID="ID",per=20,fold=3,data=srdata)
##

## End(Not run)
## Not run: 
##
mlclassKap(m=50,n=59,idSurv="OS",idEvent="event",Time="Visit",s_ID="ID",per=20,fold=3,data=srdata)
##

## End(Not run)

mlhighCox

Description

This function extracts desired number of features based on minimum log-Loss function using Cox proportional hazard model as learner method on a high dimensional survival data.

Usage

mlhighCox(cols, idSurv, idEvent, per = 20, fold = 3, data)
mlhighCox(cols, idSurv, idEvent, per = 20, fold = 3, data)

Arguments

`cols`	A numeric vector of column numbers indicating the features for which the log Loss functions are to be computed
`idSurv`	The name of the survival time variable
`idEvent`	The name of the survival event variable
`per`	Percentage of total features to be selected, default value 20
`fold`	An integer denoting number of folds in cross validation, default value 3
`data`	A data frame that contains the survival and covariate information for the subjects

Details

Performs feature Selection using Cox PH on high-dimensional data

Using the Cox proportional hazard model on the given survival data, this function selects the most significant feature based on a performance measure. The performance measure is considered as logarithmic loss function. It is defined as,

$L(f,t)=-log(f(t))$

. The features with minimum log-loss function are extracted.

Value

A dataframe containing desired number of features and the corresponding log Loss function.

Author(s)

Atanu Bhattacharjee, Gajendra K. Vishwakarma & Souvik Banerjee

References

Sonabend, R., Király, F. J., Bender, A., Bernd Bischl B. and Lang M. mlr3proba: An R Package for Machine Learning in Survival Analysis, 2021, Bioinformatics, <https://doi.org/10.1093/bioinformatics/btab039>

Examples

## Not run: 
data(hnscc)
mlhighCox(cols=c(6:15), idSurv="OS", idEvent="Death", per=20, fold = 3, data=hnscc)

## End(Not run)
## Not run: 
data(hnscc)
mlhighCox(cols=c(6:15), idSurv="OS", idEvent="Death", per=20, fold = 3, data=hnscc)

## End(Not run)

mlhighFrail

Description

This function extracts features based on minimum log-Loss function using Cox proportional hazard model as learner method on a high dimensional survival data. For those genes, we obtain frailty variances using CoxPH.

Usage

mlhighFrail(
  cols,
  idSurv,
  idEvent,
  idFrail,
  dist = "gaussian",
  per = 20,
  fold = 3,
  data
)
mlhighFrail(
  cols,
  idSurv,
  idEvent,
  idFrail,
  dist = "gaussian",
  per = 20,
  fold = 3,
  data
)

Arguments

`cols`	A numeric vector of column numbers indicating the features for which the log Loss functions are to be computed
`idSurv`	The name of the survival time variable
`idEvent`	The name of the survival event variable
`idFrail`	The name of the frailty variable
`dist`	The name of the frailty distribution. Options are "gamma", "gaussian" or "t", default is "gaussian"
`per`	Percentage of features to be selected, default value 20
`fold`	An integer denoting number of folds in cross validation, default value 3
`data`	A data frame that contains the survival and covariate information for the subjects

Details

Performs CoxPH frailty on high doimensional survival data

Using the Cox proportional hazard model on the given survival data, this function selects the most significant feature based on minimum logarithmic loss function. The logarithmic loss function is defined as,

$L(f,t)=-log(f(t))$

After selcting the most significant features, a Cox proportional hazard frailty model is fitted on the selected features. The CoxPH frailty model is defined as,

$\lambda(t)=\lambda 0(t)\nu exp{X'\beta}$

where $\nu$ is called the frailty component. The variance of the frailty term is considered as the heterogeneity among the subjects or patients. The distribution of frailty component is considered as either Gaussian, Gamma or t distribution.

Value

A dataframe containing desired number of features with corresponding frailty variances.

Author(s)

Atanu Bhattacharjee, Gajendra K. Vishwakarma & Souvik Banerjee

References

Sonabend, R., Kiraly, F. J., Bender, A., Bernd Bischl B. and Lang M. mlr3proba: An R Package for Machine Learning in Survival Analysis, 2021, Bioinformatics, <https://doi.org/10.1093/bioinformatics/btab039>

Examples

## Not run: 
data(hnscc)
mlhighFrail(cols=c(10:20), idSurv="OS", idEvent="Death", idFrail="ID", dist="gaussian",
per=20, fold = 3, data=hnscc)

## End(Not run)
## Not run: 
data(hnscc)
mlhighFrail(cols=c(10:20), idSurv="OS", idEvent="Death", idFrail="ID", dist="gaussian",
per=20, fold = 3, data=hnscc)

## End(Not run)

mlhighHet

Description

This function extracts features based on ML method, finds optimal cut-off values of features using sequencial Cox PH model and obtain the most consistent level according to the cut-offs.

Usage

mlhighHet(cols, idSurv, idEvent, idFrail, num, fold = 3, data)
mlhighHet(cols, idSurv, idEvent, idFrail, num, fold = 3, data)

Arguments

`cols`	A numeric vector of column numbers indicating the features for which the log Loss functions are to be computed
`idSurv`	The name of the survival time variable
`idEvent`	The name of the survival event variable
`idFrail`	The name of the frailty variable
`num`	Number of features to be selected
`fold`	An integer denoting number of folds in cross validation, default value 3
`data`	A data frame that contains the survival and covariate information for the subjects

Details

Performs heterogeneity analysis in gene expression

This function extracts features based on minimum log-Loss function using Cox proportional hazard model as learner method on a high dimensional survival data. For those selected genes, we obtain optimal cutoff values using minimum p-value in a Cox PH model. The Cox PH model is used sequencially for each combination of genes and all possible gene combinations are tested to obtain best possible combination with minimum BIC value. The subjects are classified according to different levels of those genes. Using a Cox PH frailty model, we obtain the most consistent level for which the frailty variance is minimum. The data is splited using cross validation technique. The performance measure is considered as logarithmic loss function. It is defined as,

$L(f,t)=-log(f(t))$

The CoxPH frailty model is defined as,

$\lambda(t)=\lambda 0(t)\nu exp{X'\beta}$

where $\nu$ is called the frailty. The variance of the frailty term is considered as the heterogeneity among the subjects or patients. Gaussian distribution with mean 0 is considered for the distribution of frailty component.

Value

dataframes containing optimal gene cutoff values and most consistent level according to those cut-offs with frailty variance.

Author(s)

Atanu Bhattacharjee, Gajendra K. Vishwakarma & Souvik Banerjee

References

Bhattacharjee, A. Vishwakarma, G.K. and Banerjee, S. A modified risk detection approach of biomarkers by frailty effect on multiple time to event data, 2020, <arXiv:2012.02102>.

Examples

## Not run: 
data(hnscc)
mlhighHet(cols=c(27:32), idSurv="OS", idEvent="Death", idFrail="ID", num=2, fold = 3, data=hnscc)

## End(Not run)
## Not run: 
data(hnscc)
mlhighHet(cols=c(27:32), idSurv="OS", idEvent="Death", idFrail="ID", num=2, fold = 3, data=hnscc)

## End(Not run)

mlhighKap

Description

This function extracts desired number of features based on minimum log-Loss function using Kaplan Meier model as learner method on a high dimensional survival data.

Usage

mlhighKap(cols, idSurv, idEvent, per = 20, fold = 3, data)
mlhighKap(cols, idSurv, idEvent, per = 20, fold = 3, data)

Arguments

`cols`	A numeric vector of column numbers indicating the features for which the log Loss functions are to be computed
`idSurv`	The name of the survival time variable
`idEvent`	The name of the survival event variable
`per`	Percentage of features to be selected, default value 20
`fold`	An integer denoting number of folds in cross validation, default value 3
`data`	A data frame that contains the survival and covariate information for the subjects

Details

Performs feature selection using Kaplan Meier method

Using the Kaplan Meier method on the given survival data, this function selects the most significant feature based on a performance measure. The performance measure is considered as logarithmic loss function. It is defined as,

$L(f,t)=-log(f(t))$

. The features with minimum log-loss function are extracted.

Value

A dataframe containing desired number of features based on minimum log Loss function

Author(s)

Atanu Bhattacharjee, Gajendra K. Vishwakarma & Souvik Banerjee

References

Sonabend, R., Kiraly, F. J., Bender, A., Bernd Bischl B. and Lang M. mlr3proba: An R Package for Machine Learning in Survival Analysis, 2021, Bioinformatics

Examples

## Not run: 
data(hnscc)
mlhighKap(cols=c(6:15), idSurv="OS", idEvent="Death", per=20, fold = 3, data=hnscc)

## End(Not run)
## Not run: 
data(hnscc)
mlhighKap(cols=c(6:15), idSurv="OS", idEvent="Death", per=20, fold = 3, data=hnscc)

## End(Not run)

High dimensional protein gene expression data

Description

High dimensional protein gene expression data

Usage

srdata
srdata

Format

A dataframe with 288 rows and 250 variables

ID: "Column/Variable name" consisting id of subjects
Visit: "Column/Variable name" consisting number of times observations recorded
event: "Column/Variable name" consisting survival event
OS: "Column/Variable name" consisting duration of overall survival
C6kine,.....,GFRalpha4: High dimensional covariates

Examples

data(srdata)
data(srdata)

Package 'highMLR'

Help Index

High dimensional head and neck cancer survival and gene expression data

Description

Usage

Format

Examples

Applications of machine learning in survival analysis by prognostic classification of genes by CoxPH model.

Description

Usage

Arguments

Value

Examples

Applications of machine learning in survival analysis by prognostic classification of genes by Kaplan-Meier estimator.

Description

Usage

Arguments

Value

Examples

mlhighCox

Description

Usage

Arguments

Details

Value

Author(s)

References

See Also

Examples

mlhighFrail

Description

Usage

Arguments

Details

Value

Author(s)

References

See Also

Examples

mlhighHet

Description

Usage

Arguments

Details

Value

Author(s)

References

See Also

Examples

mlhighKap

Description

Usage

Arguments

Details

Value

Author(s)

References

See Also

Examples

High dimensional protein gene expression data

Description

Usage

Format

Examples