| 罗燕,陈涛,王钰玺,任洪灿,高婕,吴卓璟,王晨.基于化合物结构预测人体ADME/PK性质的效能评价[J].中国药事,2023,(7):776-786 |
| 基于化合物结构预测人体ADME/PK性质的效能评价 |
| Performance Evaluation of Predicting Human ADME/PK Properties based on Compound Structure |
| 投稿时间：2022-07-04 |
| DOI：10.16153/j.1002-7777.2023.07.007 |
| 中文关键词: QSPR模型 PBPK模型 ADME/PK性质 效能评价 新药研发 |
| 英文关键词: quantitative structure property relationship model physiological based pharmacokinetic model ADME/PK properties performance evaluation new drug research & development |
| 基金项目: |
|
| 摘要点击次数: 1030 |
| 全文下载次数: 492 |
| 中文摘要: |
| 目的：整合定量结构性质关系（QSPR）模型预测化合物在人体的吸收、分布、代谢、排泄 （ADME）性质参数和基于生理药代动力学（PBPK）模型预测人体药代动力学（PK）曲线的方法，并评价该方法的预测能力。方法：以文献报道的具有体外实测理化、生物药剂学性质和临床观测PK性质的 14个化合物作为模型药物。采用ADMET Predictor软件的QSPR模型预测各个化合物的理化与生物药剂学参数，将上述预测的参数加载到GastroPlus软件的PBPK模型中预测各个化合物经口服给药后在人体的PK 曲线以及主要PK参数。对比预测与实测ADME/PK参数间的差异，以评估所用模型的预测效能。结果： QSPR模型预测的理化与生物药剂学性质参数与观测值间的绝对值较为接近，两者具有较好的线性关系（大部分参数的相关系数均接近或超过0.7）；14个化合物中，有6个化合物（43%）的最大血药浓度 （Cmax）预测值落在观测值的2倍误差范围内，9个化合物（64%）的Cmax落在观测值的3倍误差范围内； 有7个化合物（50%）的血药浓度-时间曲线下的面积（AUC）预测值落在观测值的2倍误差范围内，8个化合物（57%）的AUC落在观测值的3倍误差范围内。结论：联合QSPR和PBPK模型可用于评估化合物的ADME性质并进一步预测人体PK特征。经过当前工作的验证，表明该方法具有较高的预测能力。 |
| 英文摘要: |
| Objective: To integrate the quantitative structural property relationship (QSPR) model to predict the absorption, distribution, metabolism, excretion (ADME) property parameters of compounds in the human body and a physiological pharmacokinetic (PBPK) model to predict human pharmacokinetics (PK) profiles, and then evaluate the predictive performance of the methods. Methods: The 14 compounds with in vitro physicochemical, biopharmaceutical, and clinically observed PK properties reported in the literature were used as model drugs. The QSPR model of ADMET Predictor software was used to predict the physicochemical and biopharmaceutical parameters of each compound, and the above predicted parameters were loaded into the PBPK model of GastroPlus software to predict the PK curve and main PK parameters of each compound after oral administration. The predicted and measured ADME/PK parameters were compared to assess the predictive performance of the current models. Results: The absolute values of the physicochemical and biopharmaceutical parameters predicted by the QSPR models and the observed values were relatively close, and both of them had a good linear relationship (the correlation coeffi cients of most parameters were close to or exceeded 0.7). In these 14 compounds, there were 6 compounds (43%) whose predicted maximum plasma concentrations (Cmax) were within 2-fold error of the observed value, and 9 compounds (64%) whose Cmax were within 3-fold error; there were 7 compounds (50%) whose predicted areas under the concentration-time curve (AUC) were within the 2-fold error of the observed value, and 8 compounds (57%) whose AUC were within 3-fold error. Conclusion: The combined QSPR and PBPK models can be used to evaluate the ADME properties of compounds and further predict human PK profi les. The validation of the current work showed that the method had high predictive ability. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|