Takagi-Sugeno模型在剪接位点识别中的应用

为降低基因剪接位点识别算法复杂度和计算量,根据剪接位点上下游序列的保守特性及碱基组成随位点邻近序列GC含量变化等统计特征,建立Takagi-Sugeno模糊模型.通过模型输出值和阈值比较,判断真实的剪接位点.基于模糊似然函数的模糊聚类算法确定模型结构和前件参数,并结合最小二乘法完成该模型后件参数的识别.仿真结果表明,该算法简单,可使模糊模型的结构辨识和参数辨识同时完成,从而实现模糊模型的快速识别;能够很好地提取剪接位点附近保守序列的统计特征,为剪接位点的识别提供一种新的方法.

Application of Takagi-Sugeno model in the recognition of splice sites

To reduce the computational complexity of the recognition of splice site,a Takagi-Sugeno fuzzy model was constructed based on the feature of the conservative signal sequences and the statistical characteristics of base composition varying with GC content in up-and-down stream sequences around splice sites. A true split site was determined by comparing output and threshold value of the model.Clustering algorithm based on the fuzzy likelihood function was used to determine the structure and the premise parameters of the Takagi-Sugeno fuzzy model,which the conclusion parameters of the model was identified by least squares method.The proposed method can identify the structure and the parameters of the fuzzy model at the same time.Simulation results show the method is effective at retrieving the statistical characteristics of the conserved sequences around splicing sites.The algorithm provides a new clue for the recognition of splice site.