Threading with explicit models for evolutionary conservation of structure and sequence

We have attempted to predict the three-dimensional structures of 19 protein s for the GASPS experiment, each showing less than 25% sequence identity wi th known structures. Predictions were based on a threading method that alig ns the target sequence with the conserved cores of structural templates, as identified from structure-structure alignments of the template with homolo gous neighbors. Alternative alignments were scored using contact potentials and a position-specific score matrix derived from sequence neighbors of th e template. We find that this method identified the correct structural fami ly for 11 of the 19 targets and predicted the remaining 8 targets to be sim ilar to "none" of the templates, avoiding false positives. Threading alignm ents are relatively accurate for 10 of the 11 targets, including alignments for 6 of 7 identified at CASP3 as fold-recognition targets. These predicti ons were ranked "first place" by the GASPS assessor when compared to fold-r ecognition predictions made by other methods. It appears that threading wit h family-specific models for structure and sequence conservation has improv ed threading prediction accuracy. Published 1990 Wiley-Liss, Inc.dagger.