Department of Biochemistry and Molecular Biology, The George S. Wise Faculty of Life Sciences, Tel Aviv University, Ramat Aviv 69978, Israel.
Nucleic Acids Res. 2010 Jul;38(Web Server issue):W523-8. doi: 10.1093/nar/gkq528. Epub 2010 Jun 11.
The discrimination between functionally neutral amino acid substitutions and non-neutral mutations, affecting protein function, is very important for our understanding of diseases. The rapidly growing amounts of experimental data enable the development of computational tools to facilitate the annotation of these substitutions. Here, we describe a Random Forests-based classifier, named Mutation Detector (MuD) that utilizes structural and sequence-derived features to assess the impact of a given substitution on the protein function. In its automatic mode, MuD is comparable to alternative tools in performance. However, the uniqueness of MuD is that user-reported protein-specific structural and functional information can be added at run-time, thereby enhancing the prediction accuracy further. The MuD server, available at http://mud.tau.ac.il, assigns a reliability score to every prediction, thus offering a useful tool for the prioritization of substitutions in proteins with an available 3D structure.
区分对蛋白质功能没有影响的功能性中性氨基酸取代和影响蛋白质功能的非中性突变，对于我们理解疾病非常重要。大量不断增长的实验数据使开发计算工具成为可能，从而有助于注释这些取代。在这里，我们描述了一种基于随机森林的分类器，名为 MuD（Mutation Detector），它利用结构和序列衍生特征来评估给定取代对蛋白质功能的影响。在其自动模式下，MuD 的性能可与替代工具相媲美。然而，MuD 的独特之处在于，用户报告的特定于蛋白质的结构和功能信息可以在运行时添加，从而进一步提高预测准确性。MuD 服务器可在 http://mud.tau.ac.il 上获得，它为每个预测分配一个可靠性分数，从而为具有可用 3D 结构的蛋白质中的取代物的优先级排序提供了有用的工具。
