Combination of threading potentials and sequence profiles improves fold recognition

Using a benchmark set of structurally similar proteins, we conduct a series of threading experiments intended to identify a scoring function with an o ptimal combination of contact-potential and sequence-profile terms. The ben chmark set is selected to include many medium-difficulty fold recognition t argets, where sequence similarity is undetectable by BLAST but structural s imilarity is extensive. The contact potential is based on the log-odds of n on-local contacts involving different amino acid pairs, in native as oppose d to randomly compacted structures. The sequence profile term is that used in PSI-BLAST. We find that combination of these terms significantly improve s the success rate of fold recognition over use of either term alone, with respect to both recognition sensitivity and the accuracy of threading model s. Improvement is greatest for targets between 10 % and 20 % sequence ident ity and 60 % to 80 % superimposable residues, where the number of models cr ossing critical accuracy and significance thresholds more than doubles. We suggest that these improvements account for the successful performance of t he combined scoring function at CASP3. We discuss possible explanations as to why sequence-profile and contact-potential terms appear complementary.