Sinande's avatar
OK, I wasn't interested enough in ENCODE to do more than acknowledge its existence. Now I feel like I have to look more closely at what they mean by "function". Having read the news article in Science, I'm a bit dubious.

I'm not at all sure that things like binding TFs, being transcribed or being conserved are good indicators of function. TF binding doesn't have to mean gene expression, transcripts might be transcriptional noise destined for destruction, and a stretch of DNA can be crucially important and even have highly conserved functions while displaying a huge amount of variation (MHC genes with their hundreds of alleles, eve stripe 2 enhancer...).

(Also, did I read that right? Were most of the cell types they studied in more detail fucking *cancer cells*? :o)

I just... dunno. I hope there's actually more to it than it seemed from that article, or if there isn't, that the ENCODE people and those who'll use their data understand exactly what they have and haven't found.