摘要
基于频谱分析的基因预测方法普遍有一个缺点就是对短编码序列识别效果不佳,采用多参数Z-curve方法,再运用Fisher线性判别法对短编码序列(<100bp)进行预测,并进行了仿真验证.接着对基因突变问题采用频谱和信噪比分析方法的可行性进行了研究.
Beause oding sequene predition method based on spetrum analysis generally has a problem that the auray of identifying short oding sequene is not satisfatory. Setion 1 uses Z- Curve methods with multi-parameter and Fisher linear disriminant method to identify short oding sequene （〈lOObp）, and then omputer simulation is arried out to verify this method. The feasibility of the method using spetrum and SNR method to analyze genovariation was studied in setion 2.
出处
《数学的实践与认识》
CSCD
北大核心
2013年第14期77-84,共8页
Mathematics in Practice and Theory
