number_probes: Function to determine the number of gene probes to select for...

In nlawlor/multiClust: multiClust: An R-package for Identifying Biologically Relevant Clusters in Cancer Transcriptome Profiles

Description

Function to determine the number of gene probes to select for in the gene feature selection process.

Usage

number_probes(input, data.exp, Fixed = 1000, Percent = NULL, Poly = NULL,
  Adaptive = NULL, cutoff = NULL)

Arguments

input	String indicating the name of the file containing your gene expression matrix.
data.exp	The object containing your numeric gene expression matrix. This matrix is an output of the input_file function previously introduced in this package.
Fixed	A positive integer specifying a desired number of gene probes to select for. The default is set to 1000 gene probes.
Percent	A positive integer between 0 and 100 indicating the percentage of total gene probes to select for from the dataset.
Poly	When TRUE, a mean and variance polynomial method is used to determine the number of gene probes to select for. This method uses three second order polynomials to select for the genes with the most variable mean and standard deviations.
Adaptive	When TRUE, Gaussian mixture modeling is used to determine the number of gene probes to select.
cutoff	Positive number between 0 and 1 specifying the false discovery rate (FDR) cutoff to use with the Adaptive Gaussian mixture modeling method. The default value is set to NULL. However, when Adaptive is TRUE, cutoff should be a positive integer between 0 and 1. Common values to use are 0.05 or 0.01.

Value

Returns an object with the number of gene probes that will be selected in the gene feature selection process. If the Adaptive option is chosen, Gaussian mixture modeling files containing information about the data's mean, variance, mixing proportion, and gaussian assignment are also outputted.

Note

When using this function, the user should only use one option (Fixed, Percent, Adaptive) at a time. When using one method, all other options should be set to NULL.

This function is not needed to determine the number of gene probes to select for in the Poly gene selection method. The particular Poly method does not use a gene probe number input.

Author(s)

Peiyong Guan, Nathan Lawlor

See Also

input_file

Examples

# Example 1: Choosing a fixed gene probe number
# Load in a test file
data_file <- system.file("extdata", "GSE2034.normalized.expression.txt",
    package="multiClust")
data <- input_file(input=data_file)
gene_num <- number_probes(input=data_file, data.exp=data, Fixed=300,
    Percent=NULL, Poly=NULL, Adaptive=NULL, cutoff=NULL)

# Example 2: Choosing 50% of the total selected gene probes in a dataset
gene_num <- number_probes(input=data_file, data.exp=data, Fixed=NULL,
    Percent=50, Poly=NULL, Adaptive=NULL, cutoff=NULL)

# Example 3: Choosing the Poly method
gene_num <- number_probes(input=data_file, data.exp=data, Fixed=NULL,
    Percent=NULL, Poly=TRUE, Adaptive=NULL, cutoff=NULL)
## Not run: 
# Example 4: Choosing the Adaptive Gaussian Mixture Modeling method
# Very long computation time, so example will not be run
gene_num <- number_probes(input=data_file, data.exp=data, Fixed=NULL,
    Percent=NULL, Poly=NULL, Adaptive=TRUE, cutoff=0.01)
   
## End(Not run)

nlawlor/multiClust documentation built on May 16, 2019, 8:12 p.m.