R help - Oct 2005 - unable to compute MAD in aCGH package

Hi,

I am currently using the aCGH package in R version 2.1.0 Windows with some 
supporting packages (eg. cluster) built under R 2.1.1.Using aCGH package, 
I am able to identify regions of genomic aberrations in my cell lines 
using the HMM model. However, when I tried to use aCGH for my paraffin 
embeded tumour sample, I got the following warning.

Warning: MAD could not ben computed for one of the samples.

I had used the same R commands for both my tumour samples and cell lines. 
I had performed HMM partioning using the model AIC with a merging step 
afterwards with the threshold set at 0.25. 

Therefore, I have a few questions:

1. Could it be that my sample are 'noisy' for aCGH to compute? DNA 
extracted from paraffin embeded tissue is of a poorer quality compared to 
cell line DNA. Is there a limitation to the data distribution  itself? 

2. Or could I change the default parameters in aCGH in order to analyse my 
data?

3. A more general question, could we combine packages built under 
different versions? As mentioned above, I am using version 2.1.0 but some 
of my packages involved in aCGH are built under version 2.1.1.

Thank you.

Sincerely,
Louise 
	[[alternative HTML version deleted]]