| Objective:Screen lung adenocarcinoma for prognostic biomarkers by integrating gene expression and clinical phenotypes,and use these biomarkers to construct a prognostic predictive model of lung adenocarcinoma to identify high-risk patients and design personalized therapeutic interventions for different patients;at the same time,molecules and their functional mechanisms related to lung adenocarcinoma prognosis were initially investigated to lay the theoretical foundation for the study and application of related interventions.Methods:First,relevant data were downloaded from two databases:from the TCGA database for lung adenocarcinoma patients,including:mRNA expression levels of relevant lung adenocarcinoma genes,clinical trait data for lung adenocarcinoma patients,and clinical prognosis information for the patients;and from the Cancer Research UK database.Also from the TCGA database of lung adenocarcinoma patients.The ICGC database for lung adenocarcinoma patients includes:mRNA expression data for tumor-related genes and clinical prognosis information.Data cleaning and filtering was performed to detect the expression and prognosis of HDGF in lung adenocarcinoma patients in combination with public databases.After removing the normal samples from the cohort,the TCGA samples were divided into high and low expression groups based on the median expression values of HDGF genes,and gene function enrichment analysis was performed on the screened differential expression genes after differential expression analysis using the R language edge package.The best prognostic differentially expressed genes were screened using a one-way COX regression analysis and a LASSO-Cox regression analysis followed by a survival analysis and a mutation analysis of the screened target genes in the lung adenocarcinoma cohort.TCGA is divided into training and validation sets according to a 7:3 ratio,and the training set is used to construct a risk score prognostic model and test its efficacy.The TCGA validation set and the ICGC database lung adenocarcinoma dataset are then used as validation sets to further validate the model.Finally,after univariate and multivariate COX regression in combination with clinical characteristics related to lung adenocarcinoma prognosis,Nomogram models were constructed to predict the overall survival(OS)of lung adenocarcinoma patients and tested for their efficacy.Results:(1)HDGF was found to be highly expressed in lung adenocarcinoma with poor prognosis through public database;(2)differential gene screening and analysis were performed,and a total of 230 differential genes associated with survival of HDGF expression in lung adenocarcinoma were identified,and the upregulated genes were enriched in cortisol synthesis and secretion;(3)one-way COX regression analysis and LASSO-COX regression analysis were used to finally screen out the most robust markers associated with prognosis All nine genes were significant in survival analysis in the lung adenocarcinoma cohort and most of them had mutations in lung adenocarcinoma patients.The survival analysis showed that the overall survival time was shorter in the high-risk group than in the low-risk group,and the AUCs for 1-,3-,and 5-year survival were shown to be 0.711,0.737,and 0.773.(4)Subsequently,validation was performed with the validation set,and the AUCs for 1-,3-,and 5-year survival in the TCGA validation set were 0.725,0.613,and 0.661 and the AUCs for 1-,3-,and 5-year survival in the ICGC database validation set were 0.725,0.613,and 0.661.The AUCs for the 1-year,3-year,and 5-year survival sets are 0.627,0.673,and 0.642,indicating good model effects.(5)Single multifactor COX regression by HPRS and clinical risk factors showed that HPRS and clinical stage were independent prognostic factors for patients with lung adenocarcinoma.(6)A complex model nomogram integrating HPRS and clinical stage had good predictive performance for survival of lung adenocarcinoma patients.Conclusion:(1)Nine genetic markers(CYP17A1,KIR2DL1,CC2D2B,HHATL,SEC14L3,MUC2,MT1A,IGFBP1,RHCG),were found to be associated with lung adenocarcinoma prognosis by combining bioinformatics and statistical analysis through publicly available databases.(2)HPRS correlated with the prognosis of lung adenocarcinoma patients,and the higher the HPRS,the worse the clinical prognostic outcome of patients.(3)The individualized risk of patients was visualized and analyzed by the construction of columnar plots to help identify patients who would potentially benefit from adjuvant therapy and provide a tool for personalized treatment plans for lung adenocarcinoma patients. |
PDF Full Text Request