DOMAIN ASSIGNMENT FOR PROTEIN STRUCTURES USING A CONSENSUS APPROACH -CHARACTERIZATION AND ANALYSIS

A consensus approach for the assignment of structural domains in prote ins is presented. The approach combines a number of previously publish ed algorithms, and takes advantage of the elevated accuracy obtained w hen assignments from the individual algorithms are in agreement. The c onsensus approach is tested on a data set of 55 protein chains, for wh ich domain assignments from four automated methods were known, and for which crystallographers assignments had been reported in the literatu re. Accuracy was found to increase in this test from 72% using individ ual algorithms to 100% when all four methods were in agreement. Howeve r a consensus prediction using all four methods was only possible for 52% of the dataset. The consensus approach (using three publicly avail able domain assignment algorithms (PUU, DETECTIVE, DOMAK)) was then us ed to make domain assignments Wr a data set of 787 protein chains from the Protein Data Bank. Analysis of the assignments showed 55.7% of as signments could be made automatically, and of these, 13.5% were multi- domain proteins. Of the remaining 44.3% that could not be assigned by the consensus procedure 90.4% had their domain boundaries assigned cor rectly by at least one of the algorithms. Once identified, these domai ns were analyzed for trends in their size and secondary structure clas s. In addition, the discontinuity of each domain along the protein cha in was considered.