eighty-twenty

I didn’t make many notes on this (since it’s not really my topic) except during the question time at the end, when I caught a couple of interesting questions someone else was asking:

Q: How does it compare with homology?
A: Not well. But homology is limited too, and a hybrid approach is promising.

Comment: It seems very naïve to predict tertiary structure from secondary structure with such a simplified model - there’s a long way to go.

Q: How long were the chains you tested? How big were the errors?
A: The chains were 30 to 40 amino acids long.