Prediction of protein secondary structure at 80% accuracy

Secondary structure prediction involving up to 800 neural network predictio ns has been developed, by use of novel methods such as output expansion and a unique balloting procedure. An overall performance of 77.2%-80.2% (77.9% -80.6% mean per-chain) for three-state (helix, strand, coil) prediction was obtained when evaluated on a commonly used set of 126 protein chains. The method uses profiles made by position-specific scoring matrices as input, w hile at the output level it predicts on three consecutive residues simultan eously. The predictions arise from tenfold, cross validated training and te sting of 1032 protein sequences, using a scheme with primary structure neur al networks followed by structure filtering neural networks. With respect t o blind prediction, this work is preliminary and awaits evaluation by CASP4 . Proteins 2000;41:17-20. (C) 2000 Wiley-Liss, Inc.