This package is a developing implementation of mixOmics Bioconductor package.
The aim is to:
- implement roxygen2 documentation, and
- support MultiAssayExperiment datasets.
- support more intuitve input formats tailored to data integration
such as
formulaclass entry and support ofassaynames.
The package is not complete/tested yet, but feel free to try it using:
devtools::install_github("ajabadi/mixOmics2")You need to install the data package to be able to run the examples:
devtools::install_github("ajabadi/mixOmics.data")- August 2019:
plsfamily functions now supportMultiAssayExperimentandformulaclasses as input. check out example. - July 2019:
pcafamily functions now supportMultiAssayExperimentdata. see example.
suppressMessages({library(mixOmics2); library(mixOmics.data)})
## with X and Y as matrices
X <- linnerud$exercise
Y <- linnerud$physiological
pls.res_XY <- pls(X = X, Y = Y)
pls.res_fm <- pls(formula = Y ~ X)
pls.res_fm_MAE <- pls(data = linnerud.mae, formula = physiological~exercise)
## same output for all three
identical(pls.res_XY[-1], pls.res_fm[-1])## [1] TRUE
identical(pls.res_XY[-1], pls.res_fm_MAE[-1])## [1] TRUE
plotVar(pls.res_fm)
See examples/pls.R, examples/spls.R, examples/splsda.R,… for more.
suppressMessages({library(mixOmics2); library(mixOmics.data)})
## example with MultiAssayExperiment class
pca.res <- pca(X = multidrug.mae, assay = 'ABC.trans', ncomp = 4, scale = TRUE)
plotIndiv(pca.res, ind.names = multidrug$cell.line$Class,
group = as.numeric(as.factor(multidrug$cell.line$Class)))
See examples/pca.R, examples/spca.R, … for more.