Bioinformatics Mining Of Tumor-related Molecular Networks

One of great progresses in the study of modern oncology is to discover a lot of oncogene and tumor suppressor genes.But only the oncogene and tumor suppressor genes cannot explain all of the tumors and the problems of clinical tumors cannot be effectively solved.In fact,the tumor is a class of diseases.Every tumor contains complex molecular interactions,such as,the interactions between gene mRNA and non-coding RNA and RNA,protein and gene,etc.These complex relations altogether determines the complexity and refractoriness of the tumor.On the whole,it is necessary to grasp the ingenerate molecular mechanism of the tumor.This article first from the angle of the whole set got molecules associated with tumor.We adopted a method using literature mining software GENCLIP from PUBMED literature database unearthed literature reports of tumor associated genes;from gene expression profile chip with GE02R software database mining differentially expressed gene in the tumor samples;then from the RNAi interference database extracted those genes that have significant interference effect on tumor cell behaviors.Then we carried on the intersection of these three gene sets,the intersection and again on a set of accepted oncogene COSMIC.The last union set was the cancer related genes.For this,we analyzed the gene in molecular characteristics of the set,such as GC content,5 'UTR length,etc.We also formed a protein interaction network with the protein that the gene sets represent and analyzed the characteristics of the network,such as the degree,betweenness,etc.,Finally,with network analysis software Cytoscape identified relatively more number of links in the network hub proteins.The number of hub protein is 698.These hub proteins compared with non hub proteins,due to the relatively more connections,their being attacks,were likely to affect the stability of the entire network.Therefore,it was possible that these hub proteins identified become tumor diagnosis and treatment of the candidate targets.Secondly,the whole network was divided into local subnet that can better understand the network features.Based on the fact that,in this paper,using the gene ontology(GO)database that contains the knowledge of main biological processes,we served each of the major biological processes as functional modules in the cell.Then we got these main biological processes that have to include the associated protein,and analyzed the biological process of network that contains protein,thereby,gaining the hub proteins of each function module.The multiple function modules in the cell division,cell differentiation,programmed cell death and cell migration due to their cell phenotype was closely related to the cell fate,we called them the four phenotypic module.Through literature retrieval,we determined the phenotype of four function modules performing protein.Then,using functional modules and phenotypic protein,we analyzed the ips cells genes expression profiles.From the expression profiles,we identified the first,second and third grades downstream genes of the reprogramming factors MYC.Theses downstream genes overlap half of the "cell cycle","cell division"module genes.This result disclosed the important function of MYC for self-renewal of stem cells.Furthermore,we analyzed 10 cases liver cancer expression spectrum and got the judgment standard of each function module activated.Using the standard,we further analyzed another expression spectrum collected from precancerous lesions of liver cancer to early liver cancer and late liver cancer,decided the disease dynamic progress of each function module activated or not.Thereby,the dynamic process of the main molecular network characteristics of the diseases can be observed.Finally,we concentrated in the tumor related genes that determine a hub protein EIF4E,analyzed the EIF4E express in 55 cases of liver cancer and the adjacent tissue to cancer.At the same time,the paper analyzed the relationship between its expression and clinical data.It was found that EIF4E expression was closely related to tumor differentiation degree.