打分矩阵方法在β-发夹模体识别中的应用  

The Recognition of β-Hairpin Motifs in the Protein:Approached by Scoring Matrix

作　　者：姜雪[1,2] 胡秀珍[1] 

机构地区：[1]内蒙古工业大学理学院,呼和浩特010051 [2]沈阳工业大学辽阳校区基础部,辽阳111003

出　　处：《生物信息学》2008年第4期156-158,167,共4页Chinese Journal of Bioinformatics

基　　金：内蒙古自然科学基金资助项目200508010509

摘　　要：用打分矩阵的五种算法,从蛋白质的氨基酸序列出发,以氨基酸为参数,分别对3088个同源性小于40%的蛋白质和2878个同源性小于33%的蛋白质的β-发夹模体片断进行了识别。利用self-consistency和10-fold cross-validation两种方法对五种算法进行检验,结果表明,Claverie(1996)算法相对较好,对3088个蛋白质的平均识别精度达到了77.7%和76.1%。Based on amino acid sequences of the 3088 protein chains with identity 〈 40% and 2878 protein chains with identity 〈 33%, respectively, by using of the 5 kinds scoring matrix for predicting a - hairpin motif is proposed with parameter by amino acid in the self- consistency and 10 - fold cross - validation. Predictive results shown （1996） algorithm is significantly better than others, the overall predicted accuracies are 77.7% and 76.1% for 3088 proteins respectively.

关 键 词：Β-发夹模体 位置打分矩阵 位置打分函数 

分 类 号：Q61[生物学—生物物理学]