Ye Yuzhen, Godzik Adam
Program in Bioinformatics and Systems Biology, The Burnham Institute, 10901 N. Torrey Pines Road, La Jolla, CA 92037, USA.
Bioinformatics. 2005 May 15;21(10):2362-9. doi: 10.1093/bioinformatics/bti353. Epub 2005 Mar 3.
Existing comparisons of protein structures are not able to describe structural divergence and flexibility in the structures being compared because they focus on identifying a common invariant core and ignore parts of the structures outside this core. Understanding the structural divergence and flexibility is critical for studying the evolution of functions and specificities of proteins.
A new method of multiple protein structure alignment, POSA (Partial Order Structure Alignment), was developed using a partial order graph representation of multiple alignments. POSA has two unique features: (1) identifies and classifies regions that are conserved only in a subset of input structures and (2) allows internal rearrangements in protein structures. POSA outperforms other programs in the cases where structural flexibilities exist and provides new insights by visualizing the mosaic nature of multiple structural alignments. POSA is an ideal tool for studying the variation of protein structures within diverse structural families.
POSA is freely available for academic users on a Web server at http://fatcat.burnham.org/POSA
现有的蛋白质结构比较方法无法描述被比较结构中的结构差异和灵活性，因为它们专注于识别一个共同的不变核心，而忽略了该核心之外的结构部分。了解结构差异和灵活性对于研究蛋白质功能和特异性的演变至关重要。
使用多重比对的偏序图表示法开发了一种新的多重蛋白质结构比对方法POSA（偏序结构比对）。POSA有两个独特的特点：（1）识别并分类仅在输入结构的子集中保守的区域，（2）允许蛋白质结构内部的重排。在存在结构灵活性的情况下，POSA优于其他程序，并通过可视化多重结构比对的镶嵌性质提供了新的见解。POSA是研究不同结构家族中蛋白质结构变异的理想工具。
学术用户可通过网络服务器http://fatcat.burnham.org/POSA免费使用POSA。
