Prediction of Protein-Protein Interactions from Secondary Structures in Binding Motifs Using the Statistic Method

Protein-protein interactions and their full network are crucial to understand biological function and disease occurrence. In respect of involvement of binding motifs and specific secondary structures in protein-protein interactions, we applied a statistical method to explore the frequency with which helices, sheets and disordered secondary structures appeared on protein-protein binding motif regions and tried to predict protein-protein interactions taking this frequency as a threshold. The results have shown that i) on the average, helices and disordered structures constitute most of the binding regions (about 92%). ii)for individual binding motif the ratio may not be as significant as that in general cases. However, it is still greatly higher than that in random condition. This conclusion will be beneficial to protein-protein interaction prediction from a new orientation, secondary structures, instead of traditional ways of amino acid sequences and three-dimensional protein structures.