SAM-T04: What is new in protein-structure prediction for CASP6
SAM-T04: What is new in protein-structure prediction for CASP6
复制标题
DOI：
10.1002/prot.20730
复制
发表时间：
2005-01-01
影响因子：
2.9
通讯作者：
Hughey, R
中科院分区：
文献类型：
--
作者：
Karplus, K;Katzman, S;Hughey, R
用于预测蛋白质结构的SAM-T04方法在整个目标范围内使用单一协议，从比较建模到新折叠。该方案类似于CASP 5中使用的SAM-T02方案，但在寻找和比对模板、创建片段文库、生成蛋白质构象和对构象进行评分方面对相似序列的迭代搜索进行了改进。自动程序比简单地选择与最高得分模板的比对进行了一些改进，并且人工干预比自动程序进行了实质性改进。人类干预所做的主要改进是增加限制以构建（或保留）β折叠，以及将多结构域蛋白质拆分为单独的结构域。统一方案在整个目标难度范围内取得了中等程度的成功，但在比较建模目标方面，比CASP 6中的其他方法稍差。
The SAM-T04 method for predicting protein structures uses a single protocol across the entire range of targets, from comparative modeling to new folds. This protocol is similar to the SAM-T02 protocol used in CASP5, but has improvements in the iterative search for similar sequences in finding and aligning templates, in creating fragment libraries, in generating protein conformations, and in scoring the conformations. The automatic procedure made some improvements over simply selecting an alignment to the highest-scoring template, and human intervention made substantial improvements over the automatic procedure. The main improvements made by human intervention were from adding constraints to build (or retain) beta-sheets and from splitting multidomain proteins into separate domains. The uniform protocol was moderately successful across the entire range of target difficulty, but was somewhat less successful than other approaches in CASP6 on the comparative modeling targets.