摘要
基于CGR-游走模型和分数阶差分,用ARFIMA模型预测甲型流感病毒HA蛋白质序列。对所选取1943~2013年同源性相对较高的71条蛋白质序列,用ARFIMA(p,d,q)模型对前20个位置去拟合并且预测,除极个别外由预报区域图显示原始数据都在预报区域内,表明模型建立的比较合理,预报效果很好。这对流感病毒的研究和预测有着重要的意义。
Based on the chaos game representation walk model and the integer-order difference,the purpose of this paper is prediction for bases of influenza virus A/HA protein sequence.For the 71 selected protein sequences with high homology from 1943 to 2013,we use ARIMA(p,d,q) model to fit and predict its former 20 positions,almost all the raw data are in the forecast region except a few,showing that this model is more reasonable,and prediction of the effect is very good.It's important significance for the study and prediction of influenza virus.
出处
《食品与生物技术学报》
CAS
CSCD
北大核心
2013年第8期828-831,共4页
Journal of Food Science and Biotechnology
基金
国家自然科学基金项目(11002061)
中央高校基础研究专项项目(JUSRP21117)