sigclustTest: A statistical method for testing the significance of...

In xtsvm/CancerSubtypes: Cancer subtypes identification, validation and visualization based on multiple genomic data sets

Description

SigClust (Statistical significance of clustering) is a statistical method for testing the significance of clustering results. SigClust can be applied to assess the statistical significance of splitting a data set into two clusters. SigClust studies whether clusters are really there, using the 2-means (k = 2) clustering index as a statistic. It assesses the significance of clustering by simulation from a single null Gaussian distribution. Null Gaussian parameters are estimated from the data. Here we apply the SigClust to assess the statistical significance of pairwise subtypes. "sigclust" package should be installed.

Usage

sigclustTest(Data, group, nsim = 1000, nrep = 1, icovest = 1)

Arguments

Data	A data matrix representing the genomic data measured in a set of samples. For the matrix, the rows represent the genomic features, and the columns represents the samples.
group	The subtypes label of each sample
nsim	This is a parameter inherited from sigclust() in "sigclust" Package. Number of simulated Gaussian samples to estimate the distribution of the clustering index for the main p-value computation.
nrep	This is a parameter inherited from sigclust() in "sigclust" Package. Number of steps to use in 2-means clustering computations (default=1, chosen to optimize speed).
icovest	This is a parameter inherited from sigclust() in "sigclust" Package. Covariance estimation type: 1. Use a soft threshold method as constrained MLE (default); 2. Use sample covariance estimate (recommended when diagnostics fail); 3. Use original background noise threshold estimate (from Liu, et al, (2008)) ("hard thresholding").

Value

A matrix indicates the p-value between pairwise subtypes.

Author(s)

Xu,Taosheng [e-mail],Thuc Le [e-mail]

References

Liu, Yufeng, Hayes, David Neil, Nobel, Andrew and Marron, J. S, 2008, Statistical Significance of Clustering for High-Dimension, Low-Sample Size Data, Journal of the American Statistical Association 103(483) 1281-1293.
Huang, Hanwen, Yufeng Liu, Ming Yuan, and J. S. Marron. "Statistical Significance of Clustering Using Soft Thresholding." Journal of Computational and Graphical Statistics, no. just-accepted (2014): 00-00.

See Also

sigclust

Examples

data(GeneExp)
data(miRNAExp)
data(time)
data(status)
GBM=list(GeneExp=GeneExp,miRNAExp=miRNAExp)
result=ExecuteSNF(GBM, clusterNum=3, K=20, alpha=0.5, t=20)
group=result$group
sigclust1=sigclustTest(miRNAExp,group, nsim=500, nrep=1, icovest=3)
sigclust2=sigclustTest(miRNAExp,group, nsim=1000, nrep=1, icovest=1)

xtsvm/CancerSubtypes documentation built on May 4, 2019, 1:26 p.m.