SAAS: Short Amino Acid Sequence - A Promising Protein Secondary Structure Prediction Method of Single Sequence

Abstract

In statistical methods of predicting protein secondary structure, many researchers focus on single amino acid frequencies in α-helices, β-sheets, and so on, or the impact near amino acids on an amino acid forming a secondary structure. But the paper considers a short sequence of amino acids (3, 4, 5 or 6 amino acids) as integer, and statistics short sequence's probability forming secondary structure. Also, many researchers select low homologous sequences as statistical database. But this paper select whole PDB database. In this paper we propose a strategy to predict protein secondary structure using simple statistical method. Numerical computation shows that, short amino acids sequence as integer to statistics, which can easy see trend of short sequence forming secondary structure, and it will work well to select large statistical database (whole PDB database) without considering homologous, and Q3 accuracy is ca. 74% using this paper proposed simple statistical method, but accuracy of others statistical methods is less than 70%.

Keywords

• Short amino acid sequence
• Protein Secondary Structure Prediction
• Statistical method