Computational Biology Section Team Ranks 6th in CASP4 Competition (November/December 2000)

CASP4 represents the Fourth Community Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction. Every other year, researchers of protein structure modeling all over the world have a chance to demonstrate how good their protein-structure modeling capabilities are compared to their peers and how much progress they have made during the past two years at the CASP structure prediction competition. During each CASP prediction season, the CASP organizers will release a few dozen prediction targets (protein sequences) on the Internet, and collect the predicted 3D structures for each prediction target before a preset expiration date. Each of the prediction targets has its 3D structure already solved experimentally by an X-ray crystallography or NMR lab, but not published. All predicted structures are assessed against these experimental structures by the CASP organizers at the end of a prediction season. CASP4 started in June 2000 and ended in September 2000. Forty-three prediction targets were provided. As in previous CASPs, CASP4 had three prediction categories, ab initio folding, fold recognition, and homology modeling, which correspond to different class of prediction techniques and are applicable to different types of prediction targets. Typically a prediction team participates in one of the three categories. The ORNL team was one of 123 teams participating in the fold recognition category. The ORNL team consists of four members: Dong Xu, Oakley Crawford, Phil LoCascio, and Ying Xu (team leader), Computational Biology Section, Life Sciences Division. The main prediction tool used by the team is a threading-based structure prediction software system, called PROSPECT, which the group has been developing over the past two and one-half years. The team was ranked 6th out of the 123 teams in overall prediction performance. The team recognized more stuctural folds than any other group in the competition. (Contact: Ying Xu, 574-7263 or xuy1@ornl.gov; Funding Source: DOE KP14)