Identification of Protein Secondary Structure Based on Scoring Function
Based on scoring function of position weight matrix database for predict protein secondary structure in CB513 database in this paper.Firstly,we selected ll-residue and 21-residue fragments respectively and grouped those into three dataset (alpha,beta,Coil) according to the secondary structure of center residue;Secondly,we constructed three position weight matrix by 20 amino acids and 6 amino acids in fragments,for any one to test will be predicted the maximal one among three scores,which is calculated by the scoring function for three position weight matrix;Finally,The predicted results are corrected in the range of allowable error,and the prediction accuracy of Q3 reached 80％.
Protein secondary structurePosition weight matrixScoring functionIn error allowed scope
NETL
NSTL
万方数据
维普
