On Tuesday 17 October 2006 23:06, Di Wu wrote:
> Dear All,
>
> I have some data contain both aCGH data and affymetrix data. Same
samples
> aere used to generate the two types of arrays. aCGH arrays used are
Agilent
> custom HD-CGH microaarays(44k). I have the package aCGH in R.
>
> I want to combine the two kinds of data together according to
chromosome
> location(any other suggestion?), so that I can know the region and
gene
> list which both amplified in tumor cell and other catogorical change
> between the two types of data. I can find the way to visualize the
result
> along the chromosomes: but is there any way to calculate the results
by
> numbers and report the region or gene list in a numerical list? I
don't
> want to circle all the dot I am interested in by hand, to enlarge
them, to
> see what gene are inside.
>
> Anybody with aCGH data analysis experience? please help me out.
I would do something like the following:
1) Segment the CGH data using any of a number of CGH segmentation
methods.
The snapCGH package makes it easy to compare methods and has some
plotting
features, etc. Doing so results in (nearly) every region of the
genome
having an estimated copy number (actually log-ratio).
2) Using the data from (1), determine the copy number associated with
the
genomic locations of the affymetrix probes (yes, based on genomic
location).
Then, you have paired copy number and gene expression for each affy
probe.
With segmented CGH data and gene expression, many more analyses are
possible,
but the hypotheses to be tested need to be driven by the experiment
and
experimental design.
Sean
