Exploring Functional Gene Modules In Lung Adenocarcinoma With Weighted Gene Co-Expression Network Analysis(WGCNA)

BACKGROUND AND PURPOSE:As one of the most common cancers,lung adenocarcinoma ranks first in incidence and mortality.Researches on molecular mechanism of carcinogenesis and development help improve the early stage treatment and prognosis.Here we explored the gene expression data of lung adenocarcinoma with weighted gene co-expression network analysis,aiming at acquiring gene modules that correlate well with clinical traits and survival outcome,which enables further mining of genes in the modules.METHODS:We first established a gene co-expression network(named network A)from published RNA-seq gene expression data of 56 LUAD tumor samples with 3513 screened genes.Gene modules were extracted and correlation with clinical traits and survival outcome were assessed.Then we constructed a consensus gene co-expression network(named network B)from paired gene expression data of the 56 LUAD tumor samples and paired normal tissue samples,and the correlation between gene modules and sample type was calculated.Finally we identified the well-overlapped module pairs between network A and network B.Thereafter,we use Gene Ontology analysis and KEGG pathway analysis to explore the functional features and enriched pathways in the well-overlapped module pairs.RESULTS:With WGCNA,12 gene modules were extracted from network A.One of them was moderately correlated with tumor M grading(R2>0.4,p<0.001).The results of Cox regression denoted that four modules in network A were correlated with significantly altered hazard ratio(p<0.05).From network B,13 gene modules were established,and 9 of them were strongly correlated with pathological type.2 pairs of well-overlapped module pairs were identified,and the Gene Ontology entries and KEGG pathways enriched in these module pairs inferred functional correlation between the protein-coding genes and long non-coding genes within the same module.CONCLUSION:Gene modules determined by WGCNA were biologically meaningful.Analysis with clinical traits,survival outcomes and sample pathological type also verified the potential clinical significance of those modules.Thus,WGCNA was competent as a methodological access toward potential genetic biomarkers.