McGill Centre for Bioinformatics, McGill University, Montreal, QC, Canada.
Nucleic Acids Res. 2011 Jul;39(Web Server issue):W160-6. doi: 10.1093/nar/gkr358. Epub 2011 May 19.
RNA molecules can achieve a broad range of regulatory functions through specific structures that are in turn determined by their sequence. The prediction of mutations changing the structural properties of RNA sequences (a.k.a. deleterious mutations) is therefore useful for conducting mutagenesis experiments and synthetic biology applications. While brute force approaches can be used to analyze single-point mutations, this strategy does not scale well to multiple mutations. In this article, we present corRna a web server for predicting the multiple-point deleterious mutations in structural RNAs. corRna uses our RNAmutants framework to efficiently explore the RNA mutational landscape. It also enables users to apply search heuristics to improve the quality of the predictions. We show that corRna predictions correlate with mutagenesis experiments on the hepatitis C virus cis-acting replication element as well as match the accuracy of previous approaches on a large test-set in a much lower execution time. We illustrate these new perspectives offered by corRna by predicting five-point deleterious mutations--an insight that could not be achieved by previous methods. corRna is available at: http://corrna.cs.mcgill.ca.
RNA 分子可以通过特定的结构实现广泛的调节功能，而这些结构又取决于它们的序列。因此，预测改变 RNA 序列结构特性的突变（即有害突变）对于进行诱变实验和合成生物学应用非常有用。虽然可以使用暴力方法来分析单点突变，但这种策略不适用于多个突变。在本文中，我们提出了 corRna，这是一个用于预测结构 RNA 中多点有害突变的网络服务器。corRna 使用我们的 RNAmutants 框架来有效地探索 RNA 突变景观。它还允许用户应用搜索启发式方法来提高预测的质量。我们表明，corRna 的预测与丙型肝炎病毒顺式作用复制元件的诱变实验相关，并且在执行时间短得多的情况下，其准确性与之前方法在大型测试集上的准确性相匹配。我们通过预测五点有害突变来展示 corRna 提供的这些新视角——这是以前的方法无法实现的见解。corRna 可在：http://corrna.cs.mcgill.ca 获得。
