【摘要】 叶绿体是植物进行光合作用的主要场所,预测亚叶绿体定位对于研究其功能以及与其他大分子相互作用有重要的意义,因此更准确地预测蛋白质亚叶绿体定位成为一项必要的工作.文章建立了新的蛋白质亚叶绿体数据集,计算了氨基酸单肽分段组分信息,氨基酸二肽组分信息,预测的蛋白质二级结构信息,氨基酸指数信息,基于生物过程和分子功能的GO注释信息,以及基于PSSM矩阵的进化信息和保守信息,结合支持向量机算法(SVM)预测了亚叶绿体蛋白质定位.Jackknife检验的总体预测成功率为93.16%,同时交叉验证和独立测试也获得了较好的结果,分别为93.72%和90.65%.更多还原

【Abstract】 The chloroplast is a main place of the plant photosynthesis.It is significant to predict protein subcellular locations for understanding their functions and interactions with other molecules.So improving the predictive accuracy of subchloroplast locations becomes a necessary work.A new dataset of subchloroplast locations is constructed.And based on the calculated amino acid fragment compositional information,amino acid sequence two-peptide compositional information,predictive secondary structure of protein,amino acid index,gene ontology,evolutionary information and conservative information,the protein subchloroplast locations are predicted by using the algorithm of support vector machine.The overall prediction accuracy is 93.16%in the jackknnife test.Some better results are also obtained in the cross-validation and independent test which are 93.72% and90.65%,respectively.更多还原