| Rheumatoid arthritis is a chronic,autoimmune system disease characterized by synovial inflammation,joint swelling,autoantigen production,cartilage and bone damage.Rheumatoid arthritis and its associated complications often cause patients to gradually lose their motor function which brings serious economic burdens to families and society.The pathogenesis of rheumatoid arthritis involves many factors,and its mechanism is very complex.Although the current pathogenesis of rheumatoid arthritis is still inconclusive,related studies have shown that it is caused by both genetic and environmental factors.In order to understand the pathogenesis of rheumatoid arthritis more deeply,this article explores the genetic pathogenic factors from the perspective of gene co-expression.This article presents a new method for systematically identifying genetic etiological factors.Firstly,gene co-expression networks in normal people and rheumatoid arthritis patients were separately constructed using Weighted Gene Co-expression Network Analysis(WGCNA)algorithm.Then the mean distance hierarchical clustering algorithm combined with the dynamic hybrid tree cutalgorithm was used to detect the gene co-expression modules of normal people and patients by segmenting their gene co-expression networks respectively.Fisher’s exact test was used to verify the external reproducibility of the gene co-expression modules;the reproducibility of the co-expression modules was internally validated based on permutation test method.We used hypothesis testing method to calculate the significance of gene overlap and the significance of topology preservation in normal and patient co-expression modules based on their gene compositions and topology(density and connectivity),and we then identified specific co-expression modules for normal people and patients with rheumatoid arthritis by the thresholds.We performed functional enrichment analysis on these specific co-expression modules to find enriched GOs and Pathways using DAVID online software and clustered similar annotations using functional annotation clustering.Based on gene expression profiling samples of 20 healthy individuals and 33 rheumatoid arthritis patients,the gene co-expression networks of normal people and patients were constructed respectively,and we found 18 normal people’ gene co-expression modules and 15 rheumatoid arthritis patients’ gene co-expression modules.After reproducibility validation,16 reproducible co-expression modules of normal humans and 15 reproducible co-expression modules of rheumatoid arthritis patients were retained.Specific module analysis based on gene composition finally identified:a specific co-expression module for normal people(skyblue module)and three specific co-expression modules(brown,royalblue,and white modules)for patients;Specific module analysis based on topological change identified:a specific co-expression module for normal people(tan module)and 3 specific co-expression modules for rheumatoid arthritis patients(darkgreen,darkgrey,and green modules).Functional enrichment analysis of these specific co-expression modules identified genetically pathogenic factors that were significantly associated with rheumatoid arthritis(Appendix 1).The result of functional enrichment analysis showed that these specific co-expression modules were mainly enriched in:composition of extracellular matrix and its receptor interactions,JAK-STAT pathway,lysosomes,MAPK activity,processing of endoplasmic reticulum proteins,and T cells co-stimulation,and these enriched functions have been confirmed by relevant literature.In addition,some new pathogenic factors,such as the peptidyl-tyrosine dephosphorylation,have also been discovered in this paper,which may provide new perspectives for rheumatoid arthritis pathogenesis research.Analysis of the four co-expression modules with significantly altered topology revealed that:(1)the density of rheumatoid arthritis gene co-expression modules was denser than that of normal individuals,indicating that functions(cells Adhesion,angiogenesis,and T cell co-stimulation,etc.)enriched in these four specific co-expression modules are enhanced in patients;(2)there is not only the disappearance or appearance of co-expression relations,but also reversal co-expression relationship and change of module-dominant genes exist between normal people and rheumatoid arthritis patients,which provided a new idea for studying the pathogenesis of rheumatoid arthritis.The method proposed in this paper can systematically identify the genetic pathogenic factors associated with rheumatoid arthritis,some of which had been proven by related papers.What’s more,this article also identified a number of new pathogenic factors that are significantly associated with rheumatoid arthritis.In addition,the analysis of the topology of the specific co-expression modules also identified some of the enhanced functions in patients,and this topological analysis method also innovatively discovered some genes whose co-expression relationships were reversal and dominant status were replaced. |
