FOLD RECOGNITION AND AB-INITIO STRUCTURE PREDICTIONS USING HIDDEN MARKOV-MODELS AND BETA-STRAND PAIR POTENTIALS

Protein structure predictions were submitted for 9 of the target seque nces in the competition that ran during 1994. Targets sequences were s elected that had no known homology with any sequence of known structur e and were members of a reasonably sized family of related but diverge nt sequences. The objective was either to recognize a compatible fold for the target sequence in the database of known structures or to pred ict ab initio its rough 3D topology. The main tools used were Hidden M arkov models (HMM) for fold recognition, a beta-strand pair potential to predict beta-sheet topology, and the PHD server for secondary struc ture prediction, Compatible folds were correctly identified in a numbe r of cases and the beta-strand pair potential was shown to be a useful tool for ab initio topology prediction. (C) 1995 Wiley-Liss, Inc.