Miguel Ricardo Núñez
Department of Biochemistry, University of Cambridge, 80 Tennis Court Road, Cambridge CB2 1GA, United Kingdom.
Bioinformatics. 2004 Oct 12;20(15):2380-9. doi: 10.1093/bioinformatics/bth255. Epub 2004 Apr 8.
Most proteins have evolved to perform specific functions that are dependent on the adoption of well-defined three-dimensional (3D) structures. Specific patterns of conserved residues in amino acid sequences of divergently evolved proteins are frequently observed; these may reflect evolutionary restraints arising both from the need to maintain tertiary structure and the requirement to conserve residues more directly involved in function. Databases of such sequence patterns are valuable in identifying distant homologues, in predicting function and in the study of evolution.
A fully automated database of protein sequence patterns, Functional Protein Sequence Pattern Database (FPSPD), has been derived from the analysis of the conserved residues that are predicted to be functional in structurally aligned homologous families in the HOMSTRAD database. Environment-dependent substitution tables, evolutionary trace analysis, solvent accessibility calculations and 3D-structures were used to obtain the FPSPD. The method yielded 3584 patterns that are considered functional and 3049 patterns that are probably functional. FPSPD could be useful for assigning a protein to a homologous superfamily and thereby providing clues about function.
FPSPD is available at http://www-cryst.bioc.cam.ac.uk/~fpspd/
大多数蛋白质已经进化到执行特定功能,这些功能依赖于采用明确的三维(3D)结构。在分化进化的蛋白质的氨基酸序列中经常观察到保守残基的特定模式;这些可能反映了进化限制,其既源于维持三级结构的需要,也源于保守更直接参与功能的残基的要求。此类序列模式数据库在识别远缘同源物、预测功能和进化研究中很有价值。
一个完全自动化的蛋白质序列模式数据库,即功能蛋白质序列模式数据库(FPSPD),已从对HOMSTRAD数据库中结构比对的同源家族中预测具有功能的保守残基的分析中得出。使用环境依赖性取代表、进化追踪分析、溶剂可及性计算和三维结构来获得FPSPD。该方法产生了3584个被认为具有功能的模式和3049个可能具有功能的模式。FPSPD可用于将蛋白质归入同源超家族,从而提供有关功能的线索。