摘要
主要基于DNA序列的信噪比特征来进行基因识别.首先给出了Voss映射、Z-curve映射以及实数映射下对应的功率谱和信噪比三个公式,并用实例验证了碱基的3-周期现象与映射的选取无关.其次对于基因阈值的预测,提出了基于位置判定和基于距离平方的两种最优化方法,并对代表性生物的阈值判定结果作了分析比较.最后,提出了基于信噪比特征的基因识别算法,并预测了几种生物的基因编码区域.
In this paper, gene identification is based on DNA sequences of SNR charac- teristics.Firstly, We define the SNRs and power spectrums of different mapping,verifying the relationless of mapping and 3-Base periodicity by real example.Secondly, two kinds of opti- mization methods based on the prediction of gene threshold value are presented and analyse the results of prediction.At last,the algorithm of identifying gene founded on the SNRS feature is proposed,which predict several biological gene coding region successfully
出处
《数学的实践与认识》
CSCD
北大核心
2013年第14期53-65,共13页
Mathematics in Practice and Theory
关键词
基因预测
阈值
信噪比
gene prediction
threshold value
signal to noise ratio（SNR）
