[BioC] Package on clinical data analysis

Claus Mayer claus at bioss.ac.uk
Fri Jan 15 16:38:13 CET 2010

Hi David!

I haven't had many data sets like that but as far as I can see limma is not
only capable of comparing groups but you can also have continuous
measurements as explanatory variables, i.e. fit a multiple regression type
of model by defining the design as

Design<- model.matrix(~ var1 + var2 +..+ varn),

where var1-varn are n clinical variables. The question is, whether that is
the type of analysis you have in mind. You say "correlate" in your e-mail.
Correlation treats gene expression and clinical variable as equal, whereas
in regression you choose one of them as explanatory the other one as
response. The model above has gene expression as response (separately for
each gene). It is very well possible though that you rather want to explain
the clinical variables by the gene expression profile, i.e reverse the
roles. Limma wouldn't be the right tool for that. Or perhaps one of your
variables is survival time and you want to conduct a cox regression type of
analsyis... etc..

I think your message contains too little information to direct you to the
right tool, but if it is indeed a multiple regression you looking for, you
could still use limma (see above).


> -----Original Message-----
> From: bioconductor-bounces at stat.math.ethz.ch [mailto:bioconductor-
> bounces at stat.math.ethz.ch] On Behalf Of David martin
> Sent: 15 January 2010 10:47
> To: bioconductor at stat.math.ethz.ch
> Subject: [BioC] Package on clinical data analysis
> Hi,
> I would like to know if there is a package to analyse gene expression
> and correlate with a clinical variable ?
> Limma does all sort of gene expression analysis but is there any package
> to correlate that with clinical variables ?
> thanks,
> david
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives:
> http://news.gmane.org/gmane.science.biology.informatics.conductor

More information about the Bioconductor mailing list