Dept. of Biomedical Sciences, University of Padua, Padua, Italy.
Dept. of Engineering, Pontificia Universidad Católica del Perú, Lima, Perú.
Nucleic Acids Res. 2018 Jul 2;46(W1):W402-W407. doi: 10.1093/nar/gky360.
RepeatsDB-lite (http://protein.bio.unipd.it/repeatsdb-lite) is a web server for the prediction of repetitive structural elements and units in tandem repeat (TR) proteins. TRs are a widespread but poorly annotated class of non-globular proteins carrying heterogeneous functions. RepeatsDB-lite extends the prediction to all TR types and strongly improves the performance both in terms of computational time and accuracy over previous methods, with precision above 95% for solenoid structures. The algorithm exploits an improved TR unit library derived from the RepeatsDB database to perform an iterative structural search and assignment. The web interface provides tools for analyzing the evolutionary relationships between units and manually refine the prediction by changing unit positions and protein classification. An all-against-all structure-based sequence similarity matrix is calculated and visualized in real-time for every user edit. Reviewed predictions can be submitted to RepeatsDB for review and inclusion.
RepeatsDB-lite（http://protein.bio.unipd.it/repeatsdb-lite）是一个用于预测串联重复（TR）蛋白质中重复结构元素和单元的网络服务器。TR 是一类广泛存在但注释较差的非球状蛋白质，具有多种功能。RepeatsDB-lite 将预测扩展到所有 TR 类型，并在计算时间和准确性方面都显著优于以前的方法，对于螺旋结构的精度超过 95%。该算法利用从 RepeatsDB 数据库中提取的改进的 TR 单元库来执行迭代结构搜索和分配。网络界面提供了用于分析单元之间的进化关系的工具，并通过更改单元位置和蛋白质分类来手动改进预测。为每个用户编辑实时计算和可视化基于结构的全序列相似性矩阵。经过审查的预测可以提交给 RepeatsDB 进行审查和收录。
