论文部分内容阅读
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.
Gene identification is mainly based on the signal-noise ratio of DNA sequences.First, three formulas of power spectrum and signal-to-noise ratio under Voss mapping, Z-curve mapping and real number mapping are given, and the 3- The phenomenon of cycle is irrelevant to the selection of the mapping.Secondly, for the prediction of gene threshold, two optimization methods based on location determination and distance squared are proposed, and the threshold determination results of representative organisms are analyzed and compared.Finally, Signal-to-noise ratio feature of the gene recognition algorithm, and predict the biological coding region of several organisms.