| Gastric cancer, the world’s second highest cancer mortality, and the key factor affecting the therapeutic efficacy of gastric cancer is multi-drug resistance. Because the drug-resistance mechanism of the gastric cancer is complex and it is combined action result of complex regulatory networks involving multiple protein-coding genes, and multiple signaling pathways, intervention targeting at anyone resistance molecule is hard to reverse the drugresistance of gastric cancer. Gene expression, as a fundamental process of cellular life activities, Which can be summarized as DNA transcribed into m RNA, and then translated into protein, and transcription stage is the core step of gene expression, so the success or failure of gene expression greatly depend on the stage of transcription, and therefore it also highlights the importance of the m RNA. Our cooperation unit fourth military medical university found that there were a lot of m RNA abnormal expression in gastric resistant cells, and by measuring the expression level of these m RNA got a large number of differentially expressed genes data, and these differentially expressed genes data provide important clues for the study of multi-drug resistance of gastric cancer.However, a great number of genes are expressed specific in the drug-resistance of gastric cancer, solely using the traditional research methods that knockout and inhibit the gene one by one consume a lot of time and resources, poor feasibility. Bioinformatics can be through reasonable mathematical modeling, using powerful computing resources excavate data from vast amounts of information which are concerned by researchers and have research value, reasonable narrowing the scope of the study. Therefore, this paper through reasonable modeling and combine with network model, analyze, screen, excavate differentially expressed genes data of gastric cancer’s drug-resistance which were provided by our cooperation unit, in order to find valuable data associate with multi-drug resistance of gastric cancer. First, this paper deal with these differentially expressed genes data of gastric cancer’s drug-resistance, mainly from data redundancy and annotation errors two types of problems that exist uncertain errors or noise are processed to improve the reliability of data, after these processed data based on annotation contents are first classified, and then on the basis of initial classification, each classification will be do functional clustering by using clustering algorithm, thereby each classification data can be divided into a number of small size class. Then, on the basis of function clustering data, this paper establish each classification’s primary gene regulatory network by interactions which were excavated between genes, then measure the value of each node in each classification’s primary gene regulation network by scoring mechanism, and sort these nodes according to scores, finally according to these sorted nodes, we select appropriate core nodes, and then we design core network construction module method for these selected core nodes, next, analysis core network modules which have been excavated. After correlation analysis, this paper get the conclusion that network module composed of 20 core genes and their accessory genes are valuable data associate with multi-drug resistance of gastric cancer.On the basis of bioinformatics, from the perspective of the network model, calculation model research on regulatory networks based on gastric cancer’s drug-resistance, provides a new idea for the research of multi-drug resistance of gastric cancer. At the same time, this paper also has important theoretical and practical significance for promoting gastric cancer’s drug-resistance mechanism study, clinical treatment, and transcriptome research field, and even for other types of cancer drug resistance research also has reference. |
