氨基酸约化分类对亚线粒体蛋白定位的预测  

Prediction of Protein Submitochondria Locations Using Various Reduced Amino Acid Compositions

作　　者：张艳[1] 孙慈[1] 项新媛[1] 左永春[1] 李前忠[1] 

机构地区：[1]内蒙古大学物理科学与技术学院,呼和浩特010021

出　　处：《内蒙古大学学报（自然科学版）》2011年第3期311-317,共7页Journal of Inner Mongolia University：Natural Science Edition

基　　金：国家大学生创新性实验计划项目(No.091012603);内蒙古大学"211工程"创新人才培养项目

摘　　要：在Du和Li构建的首个亚线粒体定位数据库基础上,将线粒体蛋白依据亚线粒体位置细分为四大类进行预测.对基于氨基酸的亲疏水特征、物理化学特征和结构特征的蛋白质序列约化信息做出讨论.给出在单肽组分和六类亲疏水约化情形下的蛋白质序列最佳分割位点,结果不仅符合真实生物学现象而且范围更加精确,可为相关实验研究提供参考.提出了最佳组合参数,该参数是ω=0.10,λ=22时亲疏水残基指数值及平行相关形式的伪氨基酸组分结合全序列单肽组分,利用支持向量机算法进行预测,达到了较好的预测结果.利用本文提出的最佳组合参数对未知蛋白序列进行检验,结果显示有一定注释作用,特别是对于Inner membrane类和Matrix类的预测精度较高.The theoretical method for predicting the protein submitochondria localization was developed.Based on the database of mitochondria constructed by Du and Li,their datasets were further divided into four submitochondria localizations.The reduced amino acid compositions based on the hydropathy and physicochemical structure were discussed.The relationship between the prediction results and the length of signal-peptide was analyzed.The results discussed were consistent with the real biological phenomenon and may be helpful for the related experimental research.Based on the amino acid composition and the parallel-correlation type of hydrophobicity and hydrophilicity values（ω=0.10 and λ=22）,the SVM method performed the better prediction ability.For the independent dataset,the good prediction results were also obtained by using the combination parameters,especially for the Inner membrane class and the Matrix protein.

关 键 词：亚线粒体定位 约化组分 伪氨基酸组分 支持向量机 

分 类 号：Q61[生物学—生物物理学]