Separating normal and tumor samples where there are significant numbers of normal samples biasing correlations

[GoogleCodeExporter]

URL of the application, including the query parameters that specify the
filter:http://explorer.cancerregulome.org/all_pairs/?dataset=kirc_01oct12_A_pw&t
_label=aldh4a1

Using KIRC data set with tumor and normal samples. Looking at data table of 
results where query is Gene Expression and label is aldh4a1 in feature 1. 
Click on the first result in the data table GPD1 to display graph. enable color 
by and choose Sample Type. 
What you can see is that the correlation is being driven by the expression in 
the NT (normal) samples.
If you click on other results in the table you will see many gene associations 
are driven by the normal samples.

The goal of most queries would be to find associations in the tumor samples, 
not the normal samples. Being able to filter out normal samples and build 
correlations just within the tumor samples would avoid this problem.
Perhaps a filter to choose between sample type?

What steps will reproduce the problem?
1.
2.
3.

What is the expected output? What do you see instead?

What browser + major.minor version are you using?  What operating system
are you using?

Please provide any additional information below.

Original issue reported on code.google.com by barrysgo...@gmail.com on 2 Oct 2014 at 2:20

[GoogleCodeExporter]

RE does not currently provide this filtering ability, but as you point out it 
would be a good feature to have.
We often provide "tumor only" datasets, but not yet for KIRC.
The KIRC RE dataset is 01oct12 is also quite out of date.

We are hoping to update this dataset build before too long, and will provide a 
"tumor only" version. 

Original comment by thors...@systemsbiology.org on 7 Oct 2014 at 7:54

[GoogleCodeExporter]

Tumor only dataset for KIRC would be great. Being able to filter or toggle 
would be even better. Thanks for the reply and a great tool.

Original comment by barrysgo...@gmail.com on 8 Oct 2014 at 8:53