Using AI for Biology. https://www.bbc.com/news/science-environment-55133972
Always thankful for intelligent people and their contribution to everyone.
Printable View
Using AI for Biology. https://www.bbc.com/news/science-environment-55133972
Always thankful for intelligent people and their contribution to everyone.
This is a big deal for biologists, and it seems to be for real. The AI hype has had this result "just around the corner" for at least 30 years now. It has turned out to be a way harder problem than anyone imagined. I've just downloaded but haven't read the papers, it is going to be interesting to see how general a solution it is, whether there are classes of proteins that still defy computational solution (I'm betting yes). I'm interested to see what it can tell us about the more conformationally flexible, dynamic parts of proteins that are generally inaccessible by crystallographic methods. I'm guessing that will take even more compute power.
Some of the breakthroughs in computing technology required for better AI have taken a bit longer than we as a species would have liked. But there's no question that even today's limited abilities (which are quite amazing already) are doing good work. Virtual modeling is going to quicken a lot of things in biotech as it develops further while at the same time reducing risk, especially around "too early trials" of things.
same BBC website for this story ironically, for Nobel prize for 'DNA scissors'...
https://www.bbc.com/news/science-environment-54432589
Getting scary close to Skynet... ;)