comparing RNA-seq count data from two different sources

Question

[poisonAlien]

Hello,

So I have RNAseq read count data from two different sources.

First one is from TCGA level-3 data , which has 'raw_counts' coulmn for each gene.

Second one is from a GEO dataset, where the submitter has provided "scaled_counts" for each gene. I guess its calculated from estimateSizeFactors() from DESeq.

Now, how do I compare these count tables, one is normalized/scaled and the other one is raw ?

Do I just scale the unscaled tcga data and compare with the other one or do I have to combine both the tables and scale it before proceeding?

[Devon Ryan]

You'd be best off downloading the raw data from the GEO dataset and then processing exactly how the TCGA dataset was processed. Otherwise you're likely to just have a mess on your hands.

[poisonAlien]

Hi Devon,

You are right. I tried both method, and it does not produce expected results. Guess I will have to download raw data. Thank you.