[BioC] Tools for Combining Affy and aCGH data

[Pierre Neuvial]

Hi Tarun,

I am currently developping an R package to calculate correlations between copy number and expression measurements. Here are its main features:

- Data pre-processing: an unambiguous mapping between genome and transcriptome probes according to their position on the genome. Missing DNA copy numbers
 are inferred by taking advantage of the consistency of the copy number signal along the genome.
- Statistical data analysis: for each probe, a correlation coefficient between DNA copy number and expression data is calculated, as well as an associated p-value. These p-values are then adjusted for multiple comparisons (FDR or FWER).
- Data visualisation and interpretation: correlation coefficients and associated significances can be plotted along the genome together with cytobands; results can be exported as csv files.

As Sean mentions, our analyses show that only a subset of genes within a region of copy number amplification or deletion have their expression levels correlated with 
DNA copy number.

I hope to be able to submit this package to Bioconductor by the end of the summer. If you are interested in testing the package I will be happy to share it with you as soon as it will be stable enough to be distributed.

In case you are attending ISMB/ECCB at the end of the month, I will be presenting a poster about this work:
"GTCA: Genome Transcriptome Correlation Analysis within R" (poster number: I56)

Cheers,

Pierre.

Sean Davis wrote:
> Tarun Nayar wrote:
>> Hi,
>>  
>> Are there any tools out there to combine/ compare Affy gene expression
>> and aCGH data? A collaborator has 12 aCGH expts and 12 affy expression
>> chips from the same biological samples. He's interested in seeing if
>> regions of CN alteration correspond to differences in gene expression.
>> We've taken a stab at writing our own program for this, now I'm
>> wondering if there are any pre-made tools to do the same.
>>   
> 
> I don't know of R/bioc packages for doing this directly, but some 
> version of correlation between CGH measurements and gene expression 
> measures is probably useful.  You can use standard R commands to do 
> this.  Within any region of amplification, note that there will likely 
> be only a portion (of unknown quantity) of gene expression that is 
> correlated with copy number.
> 
> Sean
> 
> _______________________________________________
> 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
> 
> .
>