Hi,

I have pairs of transcripts that overlap on the genome, but I am not sure if they actually belong to the same gene or not. I have expression data for many different samples and I was thinking to perform a correlation test between each overlapping pair.

I can do cor.test() with method="spearman", but I guess maybe this is not the best for RNA-seq data? And also I need to correct for multiple testing.

Any suggestions on how to best perform this test? I want to remove all the gene pairs that are significantly correlated.

Thanks,

Jon

Hi Jon

I am very puzzled by your question. If you know that the transcripts originate from the same locus, in which sense could they not be from the same gene?

And why would correlation between these transcripts confirm that they belong to the same gene? How have you got your expression measure for these transcripts? If they are from the same region, then they have been quantified using (at least partially) the same reads, and so are bound to appear to correlate.

There might be some circular reasoning here.

  Simon