[Bioc-sig-seq] rna-seq: gene correlation

[Simon Anders]

Hi João

> I'm studying the correlation of genes using RNAseq. I'm wondering if you
> know some literature that could help me in that direction.. In lack of
> better solutions, do you think non parametric things like spearman would do
> the job?

In the DESeq package, we provide a function to perform a variance 
stabilizing transformation, which then gives you the possibility to use 
standard statistical techniques that require homoskedasticity. However, 
I do not have much experience on how well this works in practice for 
applications such as your.

This is because calculating correlations per gene does not make much 
sense unless you have many samples, and there are not that many 
published RNA-Seq experiments with more than a few samples. Two of the 
very few such large-scale RNA-Seq studies are the eQTL papers by 
Montomery et al. [Nature 464 (2010) 773] and by Pickrell et al. [Nature 
464 (2010) 768]. These papers in fact use Spearman correlations.

   Simon