Construction and validation of a prognostic model for lung squamous cell carcinoma based on autophagy-related genes analyzed by bioinformatics
10.3760/cma.j.cn115355-20220413-00224
- VernacularTitle:基于生物信息学分析的自噬相关基因构建肺鳞状细胞癌预后模型及验证
- Author:
Xianzhen WU
1
;
Jiaojiao LIU
;
Qi YUAN
;
Hongzhi WANG
Author Information
1. 山西省肿瘤医院 中国医学科学院肿瘤医院山西医院 山西医科大学附属肿瘤医院呼吸科，太原 030013
- Keywords:
Lung neoplasms;
Carcinoma, squamous cell;
Autophagy;
Prognosis;
Computational biology
- From:
Cancer Research and Clinic
2022;34(12):910-916
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To construct a prognostic model for lung squamous cell carcinoma (SqCLC) based on autophagy-related genes analyzed by bioinformatics and validate it.Methods:Expression profile data and clinical information of 268 SqCLC patients were downloaded from The Cancer Genome Atlas (TCGA) database and a dataset of normal lung tissues of 336 healthy people was downloaded from the Genotype Tissue Expression (GTEx) database; the autophagy-related genome was obtained from the GO_AUTOPHAGY genome of the Human Autophagy Database (HADb) and the Molecular Signature Database (MSigDB) 6.2. R 4.0.3 software was applied to analyze the differentially expressed genes between SqCLC tissues in TCGA database and normal lung tissues in GTEx database. Screening of autophagy-related genes differentially expressed between SqCLC tissues and normal lung tissues in the TCGA database (referred to as differentially expressed autophagy genes) was performed using R 4.0.3 software. The Cox proportional risk model was applied to analyze the relationship between the differentially expressed autophagy genes and prognosis of SqCLC patients in TCGA database, and a prognostic model was constructed. The SqCLC patients in TCGA database were divided into high-risk group and low-risk group based on the median risk score of the prognostic model, and the Kaplan-Meier method was used to compare the overall survival of the two groups; the time-dependent receiver operating characteristic (ROC) curve of the 3-, 5- and 10-year overall survival rates of 268 patients in TCGA database predicted by the prognostic model was plotted. Cox regression was used to analyze the independent influencing factors of overall survival of SqCLC patients in TCGA database, and the prognostic index formula was established. Based on the consistency index and restricted mean survival (RMS) curve, the predictive efficacy for the survival of patients in TCGA database between prognostic index of prognostic model risk score alone and prognostic index of risk score combined with independent influencing factors was compared. R 4.0.3 software was used to construct the nomogram for predicting patients' 3-, 5- and 10-year overall survival rates.Results:Six prognosis related differentially expressed autophagy genes were screened, and a prognostic model was constructed as: risk score=PEX14×0.337+CASPASE-8×(-0.280)+TM9SF1×0.292+UBB×0.472+P4HB×0.163+CTSA×0.173. In TCGA database, the overall survival of high-risk group was worse than that of low-risk group ( P < 0.001). Time-dependent ROC curve analysis showed that the area under the curve (AUC) of the prognostic model risk score for predicting the 3-, 5- and 10-year overall survival rates of 268 patients in TCGA database was 0.715, 0.715 and 0.831, respectively. Multivariate Cox regression analysis showed that age, staging and prognostic model risk score were independent factors affecting the overall survival of SqCLC patients in TCGA database, and the prognostic index=0.998×risk score+0.725×staging+0.559×age. The RMS curve showed that compared with the prognostic model risk score, the prognostic index combined with 3 independent prognostic factors had a better effect on predicting the overall survival (consistency index: 0.68 vs. 0.65, P =0.045). Using age, staging and prognostic model risk score, a nomogram was constructed to predict the survival of patients with SqCLC, and its calibration curve was close to the ideal curve. Conclusions:A prognostic model of SqCLC based on 6 characteristic differentially expressed autophagy-related genes has been successfully established. Internal validation shows that this model combined with other clinicopathological factors could be helpful in predicting the survival of SqCLC patients.
