A New Encoding Scheme for Protein Structure And Its Application
一种新的蛋白质结构编码方法及其应用

Based on the results of visualized clustering of tetra-peptide conformations in native protein structures, a new encoding scheme that converts 3D structure of proteins into character strings is advanced. Thus, the problem of seeking motifs of protein structures can be solved in a character sequence space. The reliability and accuracy of the scheme are validated based on two algorithms that are deve- loped for querying and discovering motifs. The concept of entropy is introduced to explore the relations between amino acid sequence and structure of proteins, and hundreds of sequence-structure motifs are obtained. Compared with results of other methods, the scheme is more accurate and explicable.