Comparing normal vs tumor using RPKM data from Wanderer/TCGA

Question

[jamie_jabalee]

Hi,

I am trying to do a rough analysis of a certain gene in various cancer types using data I downloaded from TCGA through Wanderer. The RNA seq data is provided as log2RPKM values. What I want to do is, for my gene of interest, pair up the RPKM values of normal and tumor samples from the same patient and use a simple test (for example, fold change >=|2|) to determine if the gene is up-regulated, down-regulated, or unchanged in each individual patient. The problem is that I have done some reading and am confused on whether this type of analysis makes sense with RPKM values. I am completely new to TCGA/bioinformatics so any advice is greatly appreciated!

Thanks very much in advance.

Answer 1

[Alex Reynolds]

If helpful, I wrote a TCGA expression browser that shows normalized TPM data values for ~20k genes and 31 tissues. You can select genes and tissues of interest to render box plots. The notched boxes draw a 95% CI around the median to allow comparison between normal and tumor condition and might suggest another approach. TPM measurements try to get around some of the normalization issues with RPKM data.

Answer 2

[foobarbaz]

The Wanderer tool has moved, the current URL is http://maplab.cat/wanderer/