蛋白质-RNA相互作用预测研究进展

Research progress in protein-RNA interaction prediction

蛋白质-RNA相互作用(PRI)与基因表达调控等多种生物过程密切相关。目前一般从实验和计算机预测两方面研究PRI。虽然X射线晶体衍射和核磁共振等实验方法可获得蛋白质-RNA复合物结构,但这些方法具有耗时长与花费高等缺点,并且有些蛋白质-RNA复合物结晶很难获得。特别是,随着高通量测序技术的应用,有大量的PRI需要分析,实验方法已不能满足日益增长的分析需求。为此,现已发展了4类PRI预测方法,分别为结合RNA的蛋白质残基预测、结合蛋白质的RNA小片段预测、基于序列水平的PRI预测和基于结合位点水平的PRI预测。为使有关研究人员系统了解这些预测方法,该文对上述预测方法进行了评述,并探讨其进一步发展方向。

Protein-RNA interaction（ PRI） is closely associated with many kinds of biological processes,such as gene expression regulation. Two strategies,experimental and computational identification,are generally used to study PRI. Experimental methods,such as X-ray crystal diffraction and nuclear magnetic resonance,can be used to investigate structures of protein-RNA complexes,but they are time-consuming and expensive. Furthermore,some complexes are hard to crystalize.Especially,with the application of high-throughput sequencing technology,a large number of potential PRIs are waiting to be analyzed,which has gone far beyond the capability of current experimental methods. To this end,four prediction methods have been developed for PRI,including RNA-binding protein residue prediction,protein-binding tiny RNA fragments prediction,sequence-based prediction of PRI,and binding site-based prediction of PRI. To update the related researchers on the methods of PRI prediction,and prediction models,here we reviewed these prediction models and offered some tips on how to improve related prediction methods.