Construction of Complex Features for Computational Predicting ncRNA-Protein Interaction

Non-coding RNA (ncRNA) plays important roles in many critical regulation processes. Many ncRNAs perform their regulatory functions by interacting with proteins to form RNA-protein complexes. Identifying the interaction between ncRNA and protein is very significance for understanding the function of ncRNA. In view of the cost challenge of experimental techniques, developing an accuracy computational predictive model could be of great help for the detection of the ncRNA-protein interaction. To develop an accurate prediction model, how to construct an good feature set that could characterize the interaction effectively is a critical problem. In this paper, a novel method is put forward to construct complex features for characterizing ncRNA-protein interaction (named CFRP). To obtain a comprehensive description of ncRNA-protein interaction, complex features are generated by employing non-linear transformations upon the traditional k-mer features of ncRNA and protein sequences. To reduce the dimensions of complex features, a group of discriminative features are selected by random forest. To validate the performance of the proposed method, a series of experiments are carried on several widely used public datasets. Compared with the traditional k-mer features, the CFRP complex features can boost the performance of predicting model of ncRNA-protein interaction. Meanwhile, the CFRP-based prediction model are compared with several state-of-the-art methods, and the results show that the proposed method can achieve a better performance than them. In conclusion, the complex features generated by CFRP is beneficial for building a powerful model to predict ncRNA-protein interaction.