Predicting Intrinsically Disordered Regions Based on the Structural Bias of Amino Acid Dimers
Due to many important functions of intrinsically disordered proteins, it has already become hotter and hotter research topic to distinguish intrinsically disordered regions from amino acid sequences.To accurately predict intrinsically disordered regions from amino acid sequences, a novel method was proposed to construct feature vectors based on structural bias of amino acid dimers.Compared with the frequency of amino acid monomers and dimers, the new features based on the structural bias of dimers cannot only provide the information of the components of amino acids sequence but also involve the arrangement of sequences.With the new features, BP neural network and SVM were introduced to predict intrinsically disordered regions respectively.Subsequent simulation shows improvement of predicting accuracy.It also proves the effectiveness of new features based on structural bias of amino acid dimers.
proteins intrinsic disorder structure bias prediction
Tian Feng Zhengyu Ding Fangbo Nan Yu Wang Bo He
Harbin Engineering University 145 Nantong Street, Nangang District Harbin, Heilongjiang, China, 150001
国际会议
成都
英文
46-49
2018-03-12（万方平台首次上网日期，不代表论文的发表时间）