Journal of Practical Oncology ›› 2022, Vol. 36 ›› Issue (1): 37-43.doi: 10.11904/j.issn.1002-3070.2022.01.007

• Clinical Research • Previous Articles     Next Articles

Construction of a breast cancer prognostic model based on metabolism-related genes in the TCGA database

LI Weihua1, ZHANG Guangfeng2, MA Lili3, HE Pin4, WANG Wen5, LI Dingfu1   

  1. 1. Department of Medical Imaging,First Affiliated Hospital of Shenzhen University,Shenzhen Second People's Hospital,Shenzhen 518035,China;
    2. Department of Radiology,Shenzhen Third People's Hospital;
    3. Department of Medical Imaging,Shenzhen Eye Hospital;
    4. Oncology,Chinese Academy of Medical Sciences Hospital Shenzhen Hospital;
    5. The First Affiliated Hospital of Shenzhen University,Department of Medical Pathology,Shenzhen Second People's Hospital
  • Received:2021-11-24 Revised:2022-01-06 Online:2022-02-28 Published:2022-01-21

Abstract: Objective The aims of this study were to construct a breast cancer metabolism-related prognostic model based on the TCGA database,and to reveal the relationship between metabolism and breast cancer prognosis.Methods Metabolism related genes were retrieved from KEGG database,and limma package was used to screen metabolism-related differential genes of 1109 breast cancer samples and 113 normal controls in the TCGA database.The 70% of samples were randomly selected as the training set,and the remaining 30% of the samples were used as the validation set.Univariate Cox regression was used to screen prognosis-related genes,a metabolism-related prognostic model was constructed based on glmnet package,and the model was validated by univariate and multivariate Cox regression.Results In this study,168 metabolically related differential genes were screened.After survival filtering and overfitted genes were removed,13 metabolically related genes were finally used for the construction of prognostic model,including NMNAT2、NT5E、QPRT、UGP2、MTHFD2、TSTA3、TYMP、ALDH2、ALDH1A1、IDO1、IL4I1、INPP1 and ENPP6.Survival analysis showed that the survival time of patients in the high-risk group was significantly longer than that in the low-risk group(P<0.05).Univariate and multivariate Cox regression results indicated that the risk score of the constructed metabolism related prognostic model was an independent prognostic factor for breast cancer patients.Conclusion The prognostic model constructed in this study revealed the relationship between metabolism and breast cancer prognosis,and provided a new perspective for breast cancer diagnosis and research of breast cancer.

Key words: Breast cancer, Metabolism-related genes, Prediction model, Gene signature, Prognosis

CLC Number: