Data Mining Analysis Of Prognostic Gene Biomarkers Of Metastatic Skin Cancer Based On Elastic Network

Skin cancer is a common malignant tumor throughout the world,and the rate of recurrence is relatively high,that includes metastatic tumors that occur in other tissues and metastases to the skin,thus endangering the quality of life and health of patients.Due to individual differences,the traditional treatment methods can not adapt to every patient accurately,so it is difficult to achieve the desired treatment effect for every individuals.Nowadays,with the development of gene chip,many new therapies based gene are more targeted and flexible for the treatment of skin cancer patients.Therefore,it is necessary to mine and analyze appropriate gene biomarkers according to patients' genes.Due to the high cost of gene chip technology and the large number of human genes,there are few samples of gene data and high dimensions.It is a key problem to mine effective gene biomarkers from the sample data.In the present study,the difference expression analysis was used for preliminary analysis of genes of primary tumor and metastatic samples,and proportional hazards model was used to analyse by the survival information to obtain genes associated with survival of metastatic skin cancer.Then we use the elastic network for dimensionality reduction,combined with cross-validation to dynamically shrink the regression coefficients of each gene variable,and gradually narrow down the range of gene selection,and subsequently,26 gene biomarkers were screened.A prognostic model was constructed using these 26 gene biomarkers,and the validity of the model was assessed using a training set and a verification set,which showed that the model performed well.Finally,gene function analysis of these 26 gene biomarkers was determined.Relevant studies were found to show that the genetic biomarkers identified in this paper may possess value for the follow-up clinical treatment of metastatic skin cancer.