adaptive weight protein secondary structure prediction methods applied to voting algorithm

摘要

This invention relates to protein secondary structure prediction method applying the adaptive weighted voting algorithm, a number of proteins but predicting the secondary structure of the protein to the sequence of the protein secondary structure is known as a type of protein secondary structure prediction value through the program, the secondary structure of the protein predicted by each of the program according to the corresponding ratio in the secondary structure of the known protein learning process and the process of storing and calculating the reliability indicating the confidence level of the result is calculated based on the prediction accuracy, the predicted secondary structure of the protein by using a plurality of program request when the secondary structure of the protein to the sequence of a particular protein but, by applying a weight according to the credibility of the program stored in each of the predicted results is characterized by including the weighted voting process, the process of voting. ; According to the invention, the pre-learning, but the use of the log file for the reliability of the many modules that are built by applying a different weight according to the reliability module for the results predicted by the protein by performing the voting secondary can more accurately predict the structure.