基于赖氨酸β-羟基丁酰化位点基因的五基因预后模型预测胰腺腺癌的生存和治疗反应
A five-gene prognosis model based on lysine β-hydroxybutyrylation site genes to predict the survival and therapy response in pancreatic adenocarcinoma.
作者信息
Hu Fangfang, Bai Zhibin, Yan Kai, Zhang Zheng, Zhou Jiahua
机构信息
Department of Hepatobiliary and Pancreatic Surgery, Zhongda Hospital, Medical School, Southeast University, Nanjing, Jiangsu, 210009, China.
Center of interventional Radiology and Vascular Surgery, Department of Radiology, Zhongda Hospital, Medical School, Southeast University, Nanjing, Jiangsu, 210009, China.
出版信息
Heliyon. 2024 Jul 8;10(14):e34284. doi: 10.1016/j.heliyon.2024.e34284. eCollection 2024 Jul 30.
BACKGROUND
Pancreatic adenocarcinoma (PAAD) is one of the most malignancy diseases. Lysine β-hydroxybutyrylation (Kbhb) has been reported to involve various metabolism and cancer progression.
METHODS
Data from online databases (TCGA and GEO) were retrieved for the selection of differential expressed Kbhb site genes (DTRGs). Univariate cox and LASSO analysis were performed to identify the prognostic DTRGs. Based on the optimal DTRGs, a prognostic risk score model was established. Kaplan-Meier and Receiver operator characteristic analysis were conducted to evaluate the predicting ability of the prognosis model. Generated with clinical data, independent analysis and nomogram model were performed. Finally, the differences of survival, immune cell levels, immunotherapy response, drug sensitivity between high- and low-risk groups were explored.
RESULTS
A total of 63 DTRGs were identified in PAAD, and these genes were related to cell division and apoptosis biological functions. Through univariate cox regression and LASSO analysis, 30 DTRGs were selected to be related to prognosis and five (, , , , and ) were identified as the optimal DTRGs in PAAD. Based on the five optimal DTRGs, a prognostic risk score model was constructed, with promising predictive ability in PAAD survival (AUC >0.70). High-risk group showed lower survival rate (P < 0.05). Moreover, based on the risk score, a nomogram model was also established, which possessed perfect stability. Finally, lower risk score was related to higher immune cell levels, indicating an immune activation in low-risk status, which maybe the reason for the better survival in low-risk group. Furthermore, the immunotherapy response and drug sensitivity were all higher than that in low-risk groups (P < 0.05).
CONCLUSION
A five-gene prognosis risk model which exhibit promising predictive ability in survival is constructed for patients with PAAD.
背景
胰腺腺癌（PAAD）是最恶性的疾病之一。赖氨酸β-羟基丁酰化（Kbhb）已被报道与多种代谢和癌症进展有关。
方法
从在线数据库（TCGA和GEO）检索数据，以选择差异表达的Kbhb位点基因（DTRGs）。进行单变量cox和LASSO分析以鉴定预后DTRGs。基于最佳DTRGs，建立预后风险评分模型。进行Kaplan-Meier和受试者工作特征分析以评估预后模型的预测能力。结合临床数据进行独立分析并建立列线图模型。最后，探讨高风险组和低风险组在生存、免疫细胞水平、免疫治疗反应、药物敏感性方面的差异。
结果
在PAAD中总共鉴定出63个DTRGs，这些基因与细胞分裂和凋亡生物学功能相关。通过单变量cox回归和LASSO分析，选择了30个与预后相关的DTRGs，其中5个（，，，，和）被确定为PAAD中的最佳DTRGs。基于这5个最佳DTRGs，构建了一个预后风险评分模型，对PAAD生存具有良好的预测能力（AUC>0.70）。高风险组的生存率较低（P<0.05）。此外，基于风险评分还建立了一个列线图模型，该模型具有完美的稳定性。最后，较低的风险评分与较高的免疫细胞水平相关，表明低风险状态下存在免疫激活，这可能是低风险组生存较好的原因。此外，高风险组的免疫治疗反应和药物敏感性均高于低风险组（P<0.05）。
结论
为PAAD患者构建了一个在生存方面具有良好预测能力的五基因预后风险模型。
Zotero 插件