Reconstructing Denisovan Anatomy Using DNA Methylation Maps

I can't speak to the work's accuracy (I think their PCAs are basically just the 2-3 most common facial variations seen in the wild), it was contoversial when published, but just given the facial similarity of identical twins raised in different environments, one imagines this should be very possible. https://www.pnas.org/content/early/2017/08/29/1711125114

My understanding is that it is "good accuracy" for purposes such as curiousity about what a Denisovan might have looked like. There was discussion of trying to use it for police purposes (if you have DNA of the perpetrator can you create a sketch of their appearance), and the general consensus was that it was not accurate enough for high-stakes purposes like that. But presumably, someday it may be.

This is one of those studies that is primarily "belief" based.

I for one am not a fan of these - since we can't really disprove it.

Let's predict what a Denisovan skeleton looked like based on the ancient DNA of a single specimen...

Anyway I don't mind papers like this - they are to show the public that scientists are sometimes working on problems that they can relate too.

It seems no one has been able to extract homo floresiensis DNA yet. Their warm humid environment was not conducive to DNA preservation, but the Denisova cave in Siberia has cold stable temperatures, so preserves DNA surprisingly well.