Protein's number of beta-sheets prediction using structural features

Abstract

A protein is a long one-dimensional amino acid sequence. Some subsequences of this sequence are given names and β-strand in one such subsequence. β-strands are very common and two or more such strands within one protein can form a β-sheet in the secondary structure of proteins. In its natural form, a protein is a three-dimensional entity composed of β-sheets, α-helices, and other types of substructures. Knowing the exact three-dimensional structure of a protein is the key to diagnosing several diseases and producing some drugs. In many proteins, β-sheets are the most common and the dominating substructure. Predicting the number of β-sheets in a given protein sequence is a valuable step in predicting the whole β-sheets structure in the sense that it can reduce the time complexity of the exhaustive search space examination predictors. In this research, a data-mining method for predicting the number of β-sheets is developed. The evaluations show that its performance is highly reliable.