采用粒子群优化的基因转录差异分析模型  

A Model for Analyzing Gene Transcription Difference Using Particle Swarm Optimization

作　　者：冯伟兴[1] 王科俊 

机构地区：[1]哈尔滨工程大学模式识别与智能系统研究所,哈尔滨150001

出　　处：《中国生物医学工程学报》2010年第2期229-234,共6页Chinese Journal of Biomedical Engineering

基　　金：国家高技术研究发展(863)计划(2008AA01Z148)

摘　　要：高通测序技术可以更加全面地检测实验样本中基因转录水平,这使得对样本间基因转录差异的分析精度越来越准确。根据通过DNA高通测序技术获得的基因转录区内PolⅡ蛋白个数,提出了两个样本间基因转录差异分析模型。该模型在考虑同一基因间转录差异的同时,引入了全基因的转录分布特性以提高模型分析精度。模型采用预测概率最大化统计算法进行参数求取。在概率最大化步骤中,由于难以得到模型参数解析解和提高算法效率,采用粒子群优化算法直接进行求取模型参数数值解。乳腺癌实验数据测试表明,该模型可有效分析不同样本间基因转录差异。Gene transcriptional level in an experiment can be more widely measured with high-throughput sequencing technology,which can obviously improve the analysis accuracy of gene transcriptional change between two experimental samples.Using numbers of Pol Ⅱ proteins contained inside each gene transcription region checked with DNA sequencing technology,a novel model was proposed to analyze gene transcription change considering gene transcription levels and the whole genome transcriptional distribution factor as well.Expectation-maximization（EM） statistics algorithm was utilized to resolve the model parameters.For obtaining analytical solution of model parameters and improving the efficiency of algorithm,particle swarm optimization（PSO） was used in maximization step to directly search parameters optimization numerical values.The test to analyze gene transcription change between normal and breast cancer samples verified the effectiveness of the proposed model.

关 键 词：基因 转录差异 粒子群 分析模型 

分 类 号：Q332[生物学—遗传学]