基于MIMLNN的玉米蛋白质功能预测  

Prediction of Maize Protein Function Based on MIMLNN

作　　者：陈彦明 CHEN Yan-ming(College of Information Engineering,Shanghai Maritime University,Shanghai 201306)

机构地区：[1]上海海事大学信息工程学院,上海201306

出　　处：《现代计算机》2018年第17期27-30,36,共5页Modern Computer

摘　　要：玉米作为世界上最重要的谷物之一,对其蛋白质进行预测具有重要的现实意义。为实现对玉米蛋白质的自动化功能预测,从蛋白质序列数据库提取数据并进行处理,使用优秀的多示例多标记学习算法进行玉米的蛋白质功能预测,应用主流的评价指标对预测结果进行评价,并与同类预测结果进行对比。结果显示,预测效果良好,具有一定的实用价值。As one of the world＇s most important cereals, maize has important practical significance in predicting its protein. In order to realize the auto- matic function prediction of maize protein, extracts the data from the protein sequence database and processes, and performs the protein function prediction of maize using an excellent Multi-lnstance Multi-Label learning algorithm. Uses mainstream evaluation indicators to evaluate the prediction results, and compares the results with similar experimental results. The results show that the prediction effect is good and has a certain practical value.

关 键 词：玉米 蛋白质功能预测 多示例多标记学习 机器学习 

分 类 号：S513[农业科学—作物学]