Although mathematicians and physicists have often approached protein structure prediction as an optimization problem (finding the protein conformation that minimizes the free energy), that approach has not lead to much progress over the last 50 years.
Other approaches, relying on information from other proteins, have so far been much more effective. I will present the approaches used in the "SAM-Txx" methods developed at UCSC, which have had continued success in the CASP community-wide experiments on protein structure prediction.
Copyright. All Rights Reserved.