Bioinformatics Group, Department of Computer Science, University of Freiburg, Freiburg, Germany.
Institute for Theoretical Chemistry, University of Vienna, Vienna, Austria.
Bioinformatics. 2018 Aug 1;34(15):2676-2678. doi: 10.1093/bioinformatics/bty158.
A standard method for the identification of novel RNAs or proteins is homology search via probabilistic models. One approach relies on the definition of families, which can be encoded as covariance models (CMs) or Hidden Markov Models (HMMs). While being powerful tools, their complexity makes it tedious to investigate them in their (default) tabulated form. This specifically applies to the interpretation of comparisons between multiple models as in family clans. The Covariance model visualization tools (CMV) visualize CMs or HMMs to: I) Obtain an easily interpretable representation of HMMs and CMs; II) Put them in context with the structural sequence alignments they have been created from; III) Investigate results of model comparisons and highlight regions of interest.
Source code (http://www.github.com/eggzilla/cmv), web-service (http://rna.informatik.uni-freiburg.de/CMVS).
Supplementary data are available at Bioinformatics online.
通过概率模型进行同源搜索是鉴定新 RNA 或蛋白质的标准方法。一种方法依赖于家族的定义，可以将其编码为协方差模型 (CM) 或隐马尔可夫模型 (HMM)。虽然它们是强大的工具，但由于其复杂性，以表格形式研究它们会很繁琐。这特别适用于家族群中对多个模型之间的比较的解释。协方差模型可视化工具 (CMV) 将 CM 或 HMM 可视化，以：I) 获得 HMM 和 CM 的易于解释的表示；II) 将它们与创建它们的结构序列比对联系起来；III) 研究模型比较的结果并突出显示感兴趣的区域。
源代码 (http://www.github.com/eggzilla/cmv)，网络服务 (http://rna.informatik.uni-freiburg.de/CMVS)。
补充数据可在生物信息学在线获得。
