Zhang Yi, Wang Jun, Xiao Yi
Institute of Biophysics, School of Physics, Huazhong University of Science and Technology, Wuhan 430074, Hubei, China.
Comput Struct Biotechnol J. 2020 Aug 28;18:2416-2423. doi: 10.1016/j.csbj.2020.08.017. eCollection 2020.
Most of computational methods of building RNA tertiary structure are template-based. The template-based methods usually can give more accurate 3D structures due to the use of native 3D templates, but they cannot work if the 3D templates are not available. So, a more complete library of the native 3D templates is very important for this type of methods. 3dRNA is a template-based method for building RNA tertiary structure previously proposed by us. In this paper we report improved 3D template libraries of 3dRNA by using two different schemes that give two libraries 3dRNA_Lib1 and 3dRNA_Lib2. These libraries expand the original one by nearly ten times. Benchmark shows that they can significantly increase the accuracy of 3dRNA, especially in building complex and large RNA 3D structures.
大多数构建RNA三级结构的计算方法都是基于模板的。基于模板的方法由于使用了天然三维模板，通常能够给出更准确的三维结构，但如果没有三维模板，它们就无法工作。因此，一个更完整的天然三维模板库对于这类方法非常重要。3dRNA是我们之前提出的一种基于模板的构建RNA三级结构的方法。在本文中，我们报告了通过使用两种不同方案改进的3dRNA三维模板库，得到了两个库3dRNA_Lib1和3dRNA_Lib2。这些库将原来的库扩展了近十倍。基准测试表明，它们能够显著提高3dRNA的准确性，尤其是在构建复杂和大型RNA三维结构方面。
