Study On The Mechanisms Of Lumbar Disc Degeneration Based On Text Mining And Biochip Analysis Technique

Objective:In this study, we used text mining and biochip analysis technique to analyzethe lumbar disc degeneration disease. By constructing the interaction network ofdifferential regulatory network, protein-protein interaction network and biologicpathway, we expected to demonstrate the possible mechanism of lumbar discdegeneration disease from molecular biology point of view, and provide the groundwork for making progress of lumbar disc degeneration disease in clinic.Methods：1.By using the biochip analysis technique to analyze the downloaded biochip,we obtained the differentially expressed genes (DEGs) of lumbar disc degenerationdisease. Then pathway analysis, protein-protein interaction network analysis andclustering analysis were maken.2.By using the biochip analysis technique to analyze the downloaded biochip,we obtained the differentially co-expressed genes and differentially co-expressedlinks of lumbar disc degeneration disease at differential developmental period.The obtained FDR value <0.25was considered as the threshold value ofdifferentially co-expressed genes.3.Through text mining technique analysis of lumbar disc degeneration disease,we obtained the molecular interaction relationship and screened the molecularinteraction network of mutant gene. Finally, we function enrichment analyzed theobtained network to pave the way for treating the intervertebral disc disease.Results:1. Gene expression profile of thirty-seven disc tissue samples that obtainedfrom patients with herniated discs and degenerative disc disease collected by theCooperative Human Tissue Network (CHTN) was analyzed. Differentially expressed genes between more and less degenerated discs were identified bySignificant Analysis of Microarray (SAM). Total555genes found significantlyoverexpressed in more degenerated discs with a <3%false discovery rate.Functional annotation found these genes were significantly associated withmembrane-bounded vesicle, calcium ion binding and extracellular matrix.Protein-protein interaction analysis revealed these genes might play potential keyroles in disc degeneration, including previously reported genes such as Fibronectinand β-catenin. Unsupervised clustering indicates that the widely used morphologybased on Thompson grade system has only a marginal association with molecularclassification of lumbar disc degeneration disease.2. The results of the differentially co-expressed analysis shown that there were539differentially co-expressed genes and113866couples of this co-expressedgenes have interactions. The KEGG enrichment analysis results of these obtained539differentially co-expressed genes indicated that there were10sub-pathwayshown the most significant enrichment. Compared the113866couples ofdifferentially co-expressed genes with known TF and Target Genes, we obtained62couples differentially co-expressed genes have interaction with TF and TargetGenes. It can be considered that these62couples play dominant roles in thedevelopment of intervertebral disc disease.3. The text mining analysis results indicated there were52518interactioncouples of human molecules, and there were27526interaction couples betweenhuman and mouse. Through MGI we obtained8phenotypes that relative tointervertebral disc disease, and then we screened36genes that relative to this8phenotypes. Finally, by integrating the genes with human molecular interactionnetwork, we obtained1059interaction couples of disease molecules. From the GObiological process analysis of36mutant genes, it was found that the function ofmutant genes was focused on the growth and morphological change of skeletalsystem. Conclusions:We identified the DEGs associated with progression of lumbar discdegeneration disease. Results from our study will provide the ground work for thefurther study of the molecular mechanism of lumbar disc degeneration disease.This is interesting and thus indicating that further improvement for the assessmentof progress of lumbar disc degeneration disease in clinic could be achieved. Inaddition, by screening the differentially co-expressed genes with bioinformaticsmethods, we constructed the regulatory network of these co-expressed gene. Weidentifed the target genes and biological pathways associated with intervertebraldisc disease. Finally, the text analysis was employed to identify phenotypes andgenes associated with intervertebral disc disease. Then, we constructed a networkwith these disease associated genes, and performed function annotation on thisnetwork.To sum up, based on the sreening of the differentially expressed genes andanalysis of the differentially co-expressed genes, in the combination with textmining technology to analyze molecules and pathways related with intervertebraldisc disease, we screened genes and pathways which play important roles in theprogression of lumbar disc degeneration disease disease. These research willprovide the ground work for the further study of the molecular mechanism oflumbar disc degeneration disease and for potential drug targets in the treatment ofsuch diseases. In the future, we will do further analysis on these genes andpathways combined with experiments means, to further elucidate the mechanism ofthese disease.