Protein secondary structure prediction based on the amino acids conformational classification and neural network technique

In this paper, based on the 340 protein sequences and their corresponding secondary structures got from the protein data bank (PDB), we group the 20 different amino acids into f (former), b (breaker) and n (neutral) according to their occurring frequencies in the three-state secondary structures (α-helix, β-sheets and coil), which reflect the intrinsic preference of that amino acid for a given type of secondary structure. Then we use this information to improve the protein secondary structure prediction (SSP) accuracy and get a better performance than the previous methods.