|
|
News
- August 2011
- R
package and Methods: IPCA and
sparse IPCA functions have been
implemented (as well as their
associated S3 functions). IPCA stands
for Principal Component Analysis with
Independent Loadings. It is a
combination of the advantages of both
PCA and Independent Component Analysis
(ICA). PCA is a powerful exploratory
tool if the biological question is
related to the highest variance. ICA
was recently proposed in the
literature as an alternative to PCA as
it optimizes an independence condition
that can give more meaningful
components. A preprint can be
available upon request.
- R
package and Data: The Liver
Toxicity study data has been updated
to provide geneBank IDs and gene
titles
- R
package and Data: Two other
data sets have been added: Prostate
Tumor study (gene expression) and
Metabolomic study of Yeast
(metabolomics).
 Web
interface: We are making good
progress on our associated
web-interface (now deployed on http://mixomics.qfab.org).
Few illustrative examples are also
available, and you can download the
illustrative examples and run any type
of analysis trough the interface. We
are currently developing a 'next level
analysis' to provide pathway
enrichment analyses and give the
functional annotation of the selected
genes using the iHOP database. Do not
hesitate to give us some feedback! - 'sletter:
we now have a newsletter, to
subscribe, send an email to mixomics[at]math.univ-toulouse.fr
with no subject in the body.
- June 2011
- New S3 method
network
and cim for results
from PLS model
- New code for the
valid
function to PLS-DA and SPLS-DA
models validation
- The S3 method
plot.valid
was modified to display graphical
results from valid
function for PLS-DA and SPLS-DA
models
cim and network
functions were modified to obtain
the similarity matrix in return
value- The S3 method
plotVar
was modified to obtain the
coordinates for X and Y variables in
return value
- The
predict function
has been modified to simultaneously
run either several or all prediction
methods available to predict the
classes of the test data from PLS-DA
and SPLS-DA models
- March 2011
- New function
pca and spca
are now available to perform Principal
Component Analysis (PCA) and sparse
PCA for variable selection
- The S3 methods
plotVar,
plot3dVar, plotIndiv, plot3dIndiv
were modified to generate graphical
results for pca and spca
Click
to
show/hide earlier news
- November 2010
- New
function
plot.valid
to display the results of the valid
function
- New
code for
imgCor
function for a nicer
representation of the correlation
matrices
- In
predict function the
argument 'method'
were replaced by method =
c("max.dist", "class.dist",
"centroids.dist",
"mahalanobis.dist")
- The
arguments
dendrogram,
ColSideColors and RowSideColors
were added to the cim
function
valid
function can also been performed
with missing values - Functions
pls, plsda,
spls and splsda
were modified to identify zero- or
near-zero variance predictors
- The
functions
plotVar
and plot3dVar were
modified to represent only the X
variables in the case of PLS-DA
and SPLS-DA
- The
pca function has been
improved so that the S3 methods plotIndiv,
plot3dIndiv, plotVar
and plot3dVar can be
used with these new classe
- September
2010
- Currently
improving the
pca
and nipals for
further graphical outputs
- August 2010
plsda
and splsda have been
further improved so that all the
S3 functions predict,
print, plotIndiv, plot3dIndiv
can be used with these new classes
- Several
prediction methods are now
available to predict the classes
of test data with
plsda
and splsda, see
argument 'method'
(max.dist, class.dist, centroids.dist, mahalanobis.dist)
in
the predict function
- May 2010
plsda
and splsda functions
are implemented to
perform PLS Discriminant
Analysis (PLS-DA) and sparse
PLS-DA respectively breast.tumors
data set is introduced to
illustrate the (s)PLS-DA - PCA
can also been performed with missing values using the
NIPALS algorithm and 3D plots are also
available for PCA
- Network (updated) to
display relevant associations
between variables for (r)CCA and
(s)PLS, with a new similarity
function
- A
new similarity measure has been
included in
cim
function and the arguments hclusfunc
and distfunc to display Clustered
image
maps (heatmaps)
- ... 2009
- 3D
representation to display
samples and variables for (r)CCA
- 3D
representation to display
samples and variables for (s)PLS
- The
argument
scaleY has been added
to the pls and spls functions
- (s)PLS can also
be applied when there is only 1
predictor variable
Introduction
mixOmics is an  package developed by the mixOmics team and some
collaborators. The project started in the
Institut de Mathématiques de Toulouse,
Université Paul Sabatier, Toulouse,
France.
Why
The package can be
divided into three main parts:
- Statistical methodologies to analyze
high throughput data:
- (r)CCA: (regularized) Canonical
Correlation Analysis
- (s)PLS(-DA): (sparse) Partial Least
Squares (-Discriminant Analysis)
- (s)PCA: (sparse) Principal Component
Analysis
- Several types of graphical outputs to
display the results:
- 2D plots (samples/individuals and
variables)
- Interactive 3D plots
(samples/individuals and variable)
- Interactive Relevance Networks
- Clustered Image Maps (heatmaps) with
interactive 'zoom'
- Canonical correlation scree plots
- Image maps of the correlations
matrices
- Illustrative data sets:
- breast.tumor (gene expression data)
- linnerud
- liver.toxicity (gene expression and
clinical data)
- multidrug (ABC transporters and
compounds)
- nutrimouse (gene expression and
fatty acids data)
- srbct (gene expression data)
- prostate (gene expression data)
- yeast (metabolites data)
mixOmics
can also
handle missing
values without
having to delete entire rows with missing
data.
This website gives a
full tutorial introduction to the main mixOmics R scripts
to complete the comments
and associated figures by
clicking on the symbol: 
The user can refer to
the current version of the manual 
for a full
description of each function.
The mixOmics
workflow
Download
License
The mixOmics package is released under
the GNU
Public License.
Requirements
mixOmics is a package for the R
project. You will need to have R
installed on your computer before
installing the package. For more information
about R, see the R project at http://www.r-project.org.
Download
links
Package
source, MacOS X binary, Windows binary and
reference manual are available
on
CRAN.
Installation
- Automatic installation
Start R on your computer and make sure
you are connected to the internet. You
can install the latest version
directly from the Comprehensive R
Archive Network (CRAN). At the R prompt,
type:
|
|
install.packages("mixOmics") |
|
If
you
do not have privileges on your computer
to write to the R library directory, for
example if you are using a shared unix
machine and you are not superuser, you
may need to type instead
|
|
install.packages("mixOmics",
lib = "myRlib") |
|
Instructions
for
package installation are given by typing
help(install.packages) or help(INSTALL)
at the
R prompt.
- Manual
installation
(if
you do not have Internets where R is
installed)
Download
- In the R
console, click on "Packages".
- In the
drop-down menu, select "Install
package(s) from local zip file..."
- Browse to
the zip file you would like to
install (dependencies
before mixOmics), and click
"Open".
- To load mixOmics
Type at the R prompt:
Citting mixOmics
If you are using the mixOmics package for research
that will be published or otherwise
publicly distributed, we request that you
acknowledge this with the following
citation:
Acknowledgements
The development of mixOmics is currently partially
supported by several grants:
- 2011: UQ travel grant awards for
international collaborative research,
Statistical development and analysis of
organ transplant studies from The
Prevention Of Organ Failure (PROOF,
Centre of Excellence/University of
British Columbia, Canada).
- 2011–2013: Health Research Council
Funding 11/642, Probing illness with a
novel multiomic time-course statistical
platform.
- 2011: The Maurice & Phyllis
Paykell Trust, Development of
statistical and computational methods
for the integrated 3-omics analysis of a
longitudinal study for the interaction
of obesity in acute pancreatitis.
- 2011-2013: Cooperative Research
Centres (CRC), Development of
statistical and computational methods
for the Wound
Management Innovation CRC.
Other grants that have supported, in part,
the methodological developement and
implementation of mixOmics:
- 2006-2009: Program on Food and Human
Nutrition of the French National
Research Agency, ANR PNRA 2006, project
2.23, PlastImpact
- 2006-2008: Transcriptome vs protéome,
un examen des régulations
traductionnelles par une approche de
biologie intégrative chez Lactococcus
lactis (AgroBI).
- 2005-2007: Modélisation des dynamiques
microbiennes par systèmes multi-agents
adaptatifs intégrant les données
macroscopiques et moléculaires (ANR).
- 2004-2006: Développement d'un
environnement dédié à l'analyse
statistique des données d'expression
(ACI IMPbio).
|
|
|
