p-value in paired total copy number analysis

[Yuehu...@gmail.com]

Hi,

I am wondering if there is a parameter like p-value to rank the
possibility that if there is a true variation of copy number. I did
the analysis and check report with browser, some variations are
certainly real and some are vague. If there is a function that I can
extract one or several parameters regarding this (or if it can be
included in region.xls file) it will be great.

Thanks

Yue

[Henrik Bengtsson]

Hi.

Please let us know what "analysis" you did, e.g. are you using GLAD or
CBS for segmentation?

In any case, the aroma.affymetrix package does not report any
p-values. You have to retrieve the CN model fit. For example:

fitList <- fit(cns, array=1, chromosome=22, .retResults=TRUE);
# A bit odd structure, but do str(fit) and you'll see

# Get the object returned by the DNAcopy::segment() for array 1 chromosome 22:
fit <- fitList[[1]][[22]];
print(class(fit))
[1] "DNAcopy"

Now, from there you can use the methods available in the DNAcopy
package. See its help pages for further information.

/Henrik

>
> Thanks
>
> Yue
> >
>