Song XueWu, Liu FangHao, Gao HuiEr, Yan MeiLing, Zhang FeiYu, Zhao Jia, Qin YinPeng, Li Yue, Zhang Yi
First Central Clinical College of Tianjin Medical University, Tianjin, China.
College of Computer Science, Tianjin Key Laboratory of Network and Data Security Technology, Nankai University, Tianjin, China.
Pediatr Transplant. 2023 Feb;27(1):e14379. doi: 10.1111/petr.14379. Epub 2022 Aug 30.
This study aims to establish multiple ML models and compare their performance in predicting tacrolimus concentration for infant patients who received LDLT within 3 months after transplantation.
Retrospectively collected basic information and relevant biochemical indicators of included infant patients. CMIA was used to determine tacrolimus C . PCR was used to determine the donors' and recipients' CYP3A5 genotypes. Multivariate stepwise regression analysis and stepwise elimination covariates were used for covariates selection. Thirteen machine learning algorithms were applied for the development of prediction models. APE, the ratio of the APE ≤3 ng ml and ideal rate (the proportion of the predicted value with a relative error of 30% or less) were used to evaluate the predictive performance of the model.
A total of 163 infant patients were included in this study. In the case of the optimal combination of covariates, the Ridge model had the lowest APE, 2.01 (0.85, 3.35 ng ml ). The highest ratio of the APE ≤3 ng ml was the LAR model (71.77%). And the Ridge model showed the highest ideal rate (55.05%). For the Ridge model, GRWR was the most important predictor.
Compared with other ML models, the Ridge model had good predictive performance and potential clinical application.
本研究旨在建立多个机器学习模型，并比较它们在预测移植后3个月内接受活体肝移植的婴儿患者他克莫司浓度方面的性能。
回顾性收集纳入的婴儿患者的基本信息和相关生化指标。采用化学发光微粒子免疫分析（CMIA）测定他克莫司浓度。采用聚合酶链反应（PCR）测定供体和受体的CYP3A5基因型。采用多变量逐步回归分析和逐步剔除协变量进行协变量选择。应用13种机器学习算法建立预测模型。采用绝对百分比误差（APE）、APE≤3 ng/ml的比例和理想率（相对误差≤30%的预测值比例）评估模型的预测性能。
本研究共纳入163例婴儿患者。在协变量的最佳组合情况下，岭回归模型的APE最低，为2.01（0.85，3.35 ng/ml）。APE≤3 ng/ml比例最高的是最小角回归（LAR）模型（71.77%）。岭回归模型的理想率最高（55.05%）。对于岭回归模型，肾小球滤过率加权回归（GRWR）是最重要的预测因子。
与其他机器学习模型相比，岭回归模型具有良好的预测性能和潜在的临床应用价值。
