lmdme: Linear Model on Designed Multivariate Experiments in R

This library (lmdme ) provides a flexible framework for ANOVA decomposition through linear models of multivariate designed experiments such as gene expression experiments with several experimental factors or clinical information on subjects. It allows to perform Principal Component and/or Partial Least Squares Analysis on residuals or coefficient matrices.

It implements the methodology proposed by Smilde AK et al, 2005 , Haan et al, 2007 and Nueda et al, 2007. However the available codes are very limited allowing just up to three design matrices. In this library the user can specify almost any statistical linear model through a flexible “formula” interface. It also provides statistical significance on coefficients and appropriate visualization capabilities.

The supplementary datasets can be downloaded: stemHypoxia and example2.

The following R script shows and compare results from the Nueda implementation

Example of target file. This file could be saved in tab seperated format or csv