基于Markov模型的基因组合调控模块预测

Prediction of combinatorial regulatory modules of genes based on the Markov model

研究了Markov模型在预测基因组合调控模式中的应用.首先基于基因序列特点构造Markov模型,并发展了预测基因组合调控模块的概率模型.然后结合目前基因组合调控模式数据库的特征,提出了比较模型优劣的新指标:丰度.最后利用提出的模型预测分析酵母核糖体蛋白基因中的转录调控模块,并与其它方法进行比较,结果表明Markov模型在预测基因组合调控模式时的有效性和优越性.

The paper discusses the application of the Markov chain for predicting the combinatorial regulatory modules of genes. Firstly, based on the nature of gene sequences, it constructs the Markov chain model and develops the probabilist method to predict the combinatorial regulatory motifs. Secondly, it combines the features of the com- binatorial regulatory database, and puts forward ＂enrichment＂ , a new evaluation criterion, when comparing the va- lidity of different models. Finally, the method is applied to the prediction of ribosomal protein genes of yeast. The results show the validation and superiority of the Markov model in predicting the combinatorial regulatory elements of genes through comparing several different methods.