lmdme: Linear Model on Designed Multivariate Experiments in R

The lmdme package implements analysis of variance (ANOVA) decomposition through linear models on designed multivariate experiments in R (R Development Core Team, 2012), allowing ANOVA-principal component analysis (APCA) and ANOVA-simultaneous component analysis (ASCA). It also extends both methods with the application of partial least squares (PLS) through the specification of a desired output matrix. The package is freely available on the Bioconductor website (Gentleman et al., 2004), licenced under GNU general public license. ANOVA decomposition methods for multivariate designed experiments are becoming popular in “omics” experiments (transcriptomics, metabolomics, etc.) where measurements are performed according to a predefined experimental design (Smilde et al., 2005), with several experimental factors or including subject specific clinical covariates, such as those present in current clinical genomic studies. ANOVA-PCA and ASCA are well-suited methods to study interaction patterns on multidimensional datasets. However, current R implementation of APCA is only available for Spectra data (ChemoSpec), meanwhile ASCA (Nueda et al., 2007) is based on average calculations over the indexes of up to three design matrices. Thus, no statistical inference over estimated effects is provided. Moreover, ASCA is not available in R package format. Here, R implementation on ANOVA decomposition with PCA/PLS analysis is provided. It allows a flexible formula interface the specification on almost any linear model with appropriate inference over the estimated effects and display functions for both PCA and PLS. We will present the model, implementation and a high-throughput microarray example one applied on interaction pattern analysis.