基于模糊Petri网的基因调控网络建模研究

发布时间：2018-03-14 13:09

本文选题：模糊着色Petri网　切入点：模糊Petri网　出处：《天津科技大学》2016年硕士论文　论文类型：学位论文

【摘要】：基因芯片及高吞吐量的DNA测序机技术的发展导致了超大规模的基因组数据的集成,而将这些数据转化为有价值的生物信息是目前人们面临的最大挑战,同时也成为基因调控网络研究中的一个新领域。此外研究和了解基因调控网络因果关系以及对DNA序列功能的评价对人们生活具有重要意义。相对于国外,我国在这一新兴领域的研究水平还是处于落后阶段,因此本文将从两个方面对基因调控的研究进行完善:DNA序列可信度的预测和基因表达水平影响程度的预测。针对DNA序列可信度的预测问题,将模糊Petri网(FPN)与有色Petri网(CPN)相结合,提出了一种基于模糊有色Petri网(FCPN)的方法来对DNA碱基序列可信度预测的模糊推理过程进行建模。模型中使用三个输入变量来确定DNA碱基序列的可信度,分别为height,peakness和spacing。模型中的组成部分分别对应着不同类型的模糊操作,如If—parts和Then-parts等规则。FCPN模型在模糊Petri网的基础上结合了有色Petri网,这就使FCPN方法在保证等量信息的同时,可以克服原有网络规模大、计算步骤冗余、时间长的缺点,从而提供完整的结构化表示。再此基础上建立了 DNA碱基可信度推理的FCPN模型,并得出了最终结果,与FPN方法进行比较,FCPN模型表现出较好的适应性,简化了模型的同时提高了预测的准确性。针对基因表达水平影响程度的预测问题,提出一种基于逆向推理和模糊Petri网思想的算法来对激活/抑制和目标基因间的调控关系进行建模,即根据输入的激活/抑制(activator/repressor)基因的表达水平来预测目标基因表达水平值。该方法通过FPN模型对激活/抑制和目标基因三基因的正向推理得出目标基因表达值后,采用逆向推理理论对模型进行返溯,并对结果和模型进行分析。该模型不仅可以准确的预测目标基因的表达水平,很好的模拟了基因调控网络的因果关系,而且得到了规则库中具体的作用规则以及输入基因对目标基因表达的影响程度,从而有助于制药,疾病诊断等领域的研究与探索。最后通过实例对该模型进行了验证和分析。
[Abstract]:The development of gene chip and high throughput DNA sequencer technology has led to the integration of large scale genomic data, and transforming this data into valuable biological information is the biggest challenge that people are facing. It is also a new field in the research of gene regulation network. In addition, studying and understanding the causality of gene regulation network and evaluating the function of DNA sequence are of great significance to people's life. China's research level in this emerging field is still lagging behind. Therefore, in this paper, we will improve the prediction of the reliability of DNA sequences and the degree of influence of gene expression from two aspects. In order to predict the reliability of DNA sequences, we will combine fuzzy Petri nets with colored Petri nets. A method based on fuzzy colored Petri nets (FCPN) is proposed to model the fuzzy reasoning process of DNA base sequence reliability prediction. Three input variables are used to determine the reliability of DNA base sequence. The components of the model correspond to different types of fuzzy operations, such as If-parts and Then-parts rules. The model combines colored Petri nets on the basis of fuzzy Petri nets, which makes the FCPN method guarantee the same amount of information at the same time. It can overcome the shortcomings of large scale of original network, redundant calculation steps and long time, thus providing a complete structured representation. On this basis, the FCPN model of DNA base credibility reasoning is established, and the final result is obtained. Compared with the FPN method, the FCPN model has better adaptability, simplifies the model and improves the accuracy of prediction. An algorithm based on reverse reasoning and fuzzy Petri nets is proposed to model the relationship between activation / suppression and target genes. The expression level of target gene is predicted according to the expression level of activator / inhibitor / repressor.Then the target gene expression value is obtained by positive inference of activation / inhibition and target gene by FPN model. The reverse reasoning theory is used to backtrack the model, and the results and model are analyzed. The model can not only accurately predict the expression level of target gene, but also simulate the causality of gene regulation network. Moreover, the specific action rules in the rule base and the influence of input genes on the expression of target genes are obtained, which is helpful for the research and exploration of pharmaceutical and disease diagnosis fields. Finally, the model is verified and analyzed by an example.
【学位授予单位】：天津科技大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：Q811.4;TP301.1

【相似文献】