New method for accurate prediction of solvent accessibility from protein sequence

A novel method was developed for predicting the solvent accessibility. Base d on single sequence data, this method achieved 71.5% accuracy with a corre lation coefficient of 0.42 in a database of 704 proteins with threshold of 20% for a two-state-defining solvent accessibility. Prediction in a data su bset of 341 monomeric proteins achieved 72.7% accuracy with a correlation c oefficient of 0.43. On the average, prediction over short chains gives bett er results than that over long chains. With a solvent accessibility thresho ld of 20%, prediction over 236 monomeric proteins with chain length < 300 a mino acid residues achieved 75.3% accuracy with a correlation coefficient o f 0.44 by jackknife analysis, which is higher than that obtained by previou s methods using multiple sequence alignments. Proteins 2001;42:1-5. (C) 200 0 Wiley-Liss, Inc.