Entering edit mode
Michael Breen
▴
370
@michael-breen-5999
Last seen 9.7 years ago
Hi all,
We are interested in using VirtualArray while mining a few datasets
into
order to ask more theoretical questions than the data was originally
intended for. However, to set the foundation I need to merge various
data-sets from very similar platforms.
I have just a couple questions regarding the application of
VirtualArray.
To reduce complexity in our dialogue I will just refer to 2 GEO
datasets
for the time being.
First, we have checked the expression summaries of each of the
GeoSets. How
does your program infer which normalizing to perform on one data-set
vs.
the other to make the distributions more similar to each other? I can
make
assumptions with the below example, however do you have a test for
this?
> summary(Eset1)[,1:3]
Pat1 Pat2 Pat3
"Min. :-1.48434 " "Min. :-1.628659 " "Min. :-1.63216 "
"1st Qu.:-0.05553 " "1st Qu.:-0.075109 " "1st Qu.:-0.08433 "
"Median : 0.01388 " "Median :-0.001093 " "Median : 0.02052 "
"Mean : 0.01210 " "Mean : 0.002367 " "Mean : 0.02474 "
"3rd Qu.: 0.10731 " "3rd Qu.: 0.075719 " "3rd Qu.: 0.16198 "
"Max. : 2.01628 " "Max. : 2.439713 " "Max. : 1.62180 "
> summary(Eset2)[,1:3]
Pat1 Pat2 Pat3
"Min. : 0.8523 " "Min. : 0.8373 " "Min. : 0.6775 "
"1st Qu.: 4.6027 " "1st Qu.: 4.4968 " "1st Qu.: 4.4723 "
"Median : 6.0780 " "Median : 6.0009 " "Median : 5.9944 "
"Mean : 6.1098 " "Mean : 6.0757 " "Mean : 6.0891 "
"3rd Qu.: 7.4339 " "3rd Qu.: 7.4802 " "3rd Qu.: 7.5196 "
"Max. :13.5664 " "Max. :13.4372 " "Max. :13.3803 "
Second, when compiling the new expression set with:
> my_virtualArrays$iPSC_hESC_noBatchEffect <-
virtualArrayExpressionSets()
While this runs we have run into 2 problems. The first being that our
annotation type is GPL5175 for both Esets in this example. And we will
receive the following error: Error in eval(expr, envir, enclos) :
object 'GPL5175SYMBOL' not found.
The second is that this step also relies upon the BiocParallel
bioconductor
source package which does not seem to be installing for R v. 3.0.2. I
have
attempted bioconductor download, install.packages command,
reinstallation
of R etc... with no advances.
Any insight to the above questions would be great.
Yours,
Michael
[[alternative HTML version deleted]]
