基于压缩感知预测凋亡蛋白亚细胞位点  

Prediction of apoptosis protein subcellular location based on compressive sensing

作　　者：石雪娜[1] 王瑞平[1] 

机构地区：[1]北京交通大学计算机与信息技术学院,北京100044

出　　处：《北京生物医学工程》2015年第1期70-74,共5页Beijing Biomedical Engineering

基　　金：北京交通大学基本科研业务费专项基金(2011JBM021)资助

摘　　要：目的蛋白质的亚细胞位点信息有助于了解蛋白质的功能,同时还可以为新药物的研发提供帮助。方法在本研究中,采用压缩感知算法对凋亡蛋白的亚细胞位点进行分类。结果经过在2个常用的数据集上使用Jackknife测试,在ZD98数据集上取得了90.6%的预测准确率,在ZW225数据集上取得了87.3%的准确率。结论实验结果表明压缩感知算法对已凋亡蛋白的亚细胞定位研究效果良好,且避免了特征提取的过程。Objective The information of protein subcellular location helps us to understand protein function and to design new drugs. Methods In this paper, compressive sensing algorithm is adopted to classify protein subcellular location. Results By Jackknife test on datasets, our overall accuracies for ZD98 and ZW225 are 90.6% and 87.3% respectively. Conclusions The effect of compressive sensing algorithm is significant to classify apoptosis protein subcellular location. And compressive sensing can avoid the process of feature extraction.

关 键 词：凋亡蛋白 亚细胞位点 压缩感知 

分 类 号：R318[医药卫生—生物医学工程] Q51[医药卫生—基础医学]