Objective: Programmed cell death (PCD) has therapeutic potential for a variety of malignant tumors, including lung cancer. In this study, we used PCD and bioinformatics to construct a prognostic model for lung cancer and explore new therapeutic strategies. Methods: Multiple bioinformatics algorithms (co-expression analysis, univariate Cox’s analysis, multivariate Cox’s analysis, and cross-validation) were used to screen PCD-related genes and construct a risk model. Lung cancer patients were divided into training and testing groups in a ratio of 7:3. The prognostic model was validated by comparing the risk scores of the high-risk and low-risk groups using receiver operating characteristic (ROC) curves, nomograms, and independent prognostic analyses. In addition, PCD patterns were classified and compared in terms of survival time, immune microenvironment and pathway regulation using consensus clustering methods with the validation of principal component analysis (PCA). Single-cell RNA sequencing (scRNA-seq) analysis was applied to validate screened PCD-related genes in this risk model. Results: Twelve risk genes were identified, including BIK, CDCP1, CHEK2, FADD, GLS2, IL33, ITPRIP, KRT8, MELK, MMP9, PTGIS and TRIB3, to construct prognostic risk model. In the process of lung cancer, the most significantly up-regulated gene was TRIML2, and the most down-regulated gene was GLS2. ROC curves, nomograms, and independent prognostic analyses confirmed the accuracy of risk model to predict the prognosis of lung cancer, indicating that it can be regarded as an independent prognostic model. In the immune cell infiltration, we found that patients with an increased M0 macrophage had a poorer prognosis. Drug sensitivity testing after reliable risk modeling identified three molecularly targeted drugs for lung cancer patients in the high-risk group, namely, Staurosporine, Luminespib and Docetaxel. scRNA-seq results further analyzed the reliability of ITPRIP and KRT8 as prognostic targets. Conclusion: This study identified twelve PCD-related genes and constructed an accurate risk model based on bioinformatics analysis, which can be used for prognostic prediction and design of clinical treatment strategies.