基于结构的单链和双链DNA结合蛋白的分类研究

Protein-DNA interactions are essential for many biological processes. However, the structural mechanisms underlying these interactions are not fully understood. We introduce a new motif called protein tunnel features and OB-fold structural alignment (OSA) to study the difference between double-stranded DNA-binding proteins (DSBs) and single-stranded DNA binding proteins (SSBs). Based the incorporative features were constructed this classifiers with support vector machine (SVM). The predictor was evaluated using a 10-fold cross validation in HOLO (36/218) and APO (36/46) non-redundant datasets. The best SVM classifier results achieved by our proposed method for: accuracy 83.23%, sensitivity 83.76%, specificity 79.75%, AUC 0.8842 and applied to the independent test; Moreover, HOLO and APO dataset were used to the independent test, and analyzed HOLO-APO proteins conformational change with DNA-binding. The results show that the properties of the surfaces bottleneck radius, tunnel length, and tunnel curvature conformational features can be applicable for de novo protein-function prediction. In conclusion, we present a novel methodology to characterize protein surface, which can accurately tell apart SSBs from DSBs. The strength of our method in recognizing fine-tuned differences on DNA binding interfaces make it applicable for many other molecular recognition problems, with potential implications for drug design.