isva:
Independent Surrogate Variable Analysis
Description
Independent Surrogate Variable Analysis is an algorithm for feature
selection in the presence of potential confounding factors, specially
designed for the analysis of large-scale high-dimensional quantitative
genomic data (e.g microarrays). It uses Independent Component Analysis
(ICA) to model the confounding factors as independent surrogate
variables (ISVs). These ISVs are included as covariates in a
multivariate regression model to subsequently identify features that
correlate with a phenotype of interest independently of these
confounders. Two ICA implementations are offered: JADE from the JADE
R-package and fastICA from the fastICA R-package.
Details
Package: |
isva |
Type: |
Package |
Version: |
1.9 |
Date: |
2017-01-13 |
License: |
GPL-2 |
LazyLoad: |
yes |
There are two internal functions. One function (EstDimRMT) performs the dimensionality estimation using a Random Matrix Theory approximation. The other function (isvaFn) is the main engine function and performs the modelling of confounding factors using Independent Component Analysis (ICA). Briefly, ICA is applied on the residual variation orthogonal to that of the phenotype of interest. DoISVA is the main user function, performing feature selection using the constructed independent surrogate variables as covariates.
References
Independent Surrogate Variable Analysis to deconvolve confounding factors in large-scale microarray profiling studies. Teschendorff AE, Zhuang JJ, Widschwendter M. Bioinformatics. 2011 Jun 1;27(11):1496-505.