Laboratory of Systems Biology and Bioinformatics, Institute of Biochemistry and Biophysics and COE in Biomathematics, University of Tehran, Tehran, Iran.
J Theor Biol. 2012 Jan 21;293:55-64. doi: 10.1016/j.jtbi.2011.09.033. Epub 2011 Oct 10.
Cellular functions are mediated by various biological processes including biomolecular interactions, such as protein-protein, DNA-protein and RNA-protein interactions in which RNA-Protein interactions are indispensable for many biological processes like cell development and viral replication. Unlike the protein-protein and protein-DNA interactions, accurate mechanisms and structures of the RNA-Protein complexes are not fully understood. A large amount of theoretical evidence have shown during the past several years that computational geometry is the first pace in understanding the binding profiles and plays a key role in the study of intricate biological structures, interactions and complexes. In this paper, RNA-Protein interaction interface surface is computed via the weighted Voronoi diagram of atoms. Using two filter operations provides a natural definition for interface atoms as classic methods. Unbounded parts of Voronoi facets that are far from the complex are trimmed using modified convex hull of atom centers. This algorithm is implemented to a database with different RNA-Protein complexes extracted from Protein Data Bank (PDB). Afterward, the features of interfaces have been computed and compared with classic method. The results show high correlation coefficients between interface size in the Voronoi model and the classical model based on solvent accessibility, as well as high accuracy and precision in comparison to classical model.
细胞功能是由各种生物过程介导的，包括生物分子相互作用，如蛋白质-蛋白质、DNA-蛋白质和 RNA-蛋白质相互作用，其中 RNA-蛋白质相互作用对于许多生物过程是必不可少的，如细胞发育和病毒复制。与蛋白质-蛋白质和蛋白质-DNA 相互作用不同，RNA-蛋白质复合物的准确机制和结构尚未完全理解。近年来，大量理论证据表明，计算几何是理解结合谱的第一步，在研究复杂的生物结构、相互作用和复合物方面起着关键作用。在本文中，通过原子的加权 Voronoi 图计算 RNA-蛋白质相互作用界面。使用两种过滤操作为界面原子提供了经典方法的自然定义。使用修改后的原子中心凸包修剪远离复合物的 Voronoi 面的无界部分。该算法已应用于从蛋白质数据库（PDB）中提取的不同 RNA-蛋白质复合物的数据库。然后，计算了界面的特征，并与经典方法进行了比较。结果表明，Voronoi 模型中的界面大小与基于溶剂可及性的经典模型之间具有很高的相关系数，并且与经典模型相比具有很高的准确性和精度。
