当前位置: X-MOL 学术Pestic. Biochem. Phys. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
A revised model of fungicide translaminar activity
Pesticide Biochemistry and Physiology ( IF 4.7 ) Pub Date : 2020-05-01 , DOI: 10.1016/j.pestbp.2020.104597
Carla J R Klittich 1 , Nick X Wang 1 , Yu Zhang 1 , L Boyd Rowland 1
Affiliation  

Translaminar redistribution is valuable for fungicide activity but difficult to measure and predict. The translaminar activity of 38 fungicides active against cucumber powdery mildew was measured experimentally and used to develop a QSAR (Quantitative structure-activity relationship) model of translaminar movement from calculated parameters. Over 300 physiochemical parameters generated from energy-minimized 3D structures were considered and one-parameter, two-parameter, and five-parameter models were developed. The one-parameter lipophilicity model explained 39% of variability in translaminar activity in the full dataset but none of the variability in the small succinate dehydrogenase inhibitor (SDHI) set. Adding a polar surface area parameter to the lipophilicity parameter improved predictability to 52% and explained over 70% of the variability in the SDHI class. The expanded model with five physiochemical parameters explained more than 80% of the variability in overall translaminar redistribution. The three additional parameters were correlated with molecular size and reactivity. The models were validated with a Leave-One-Out method that showed excellent robustness (r2adj = 0.83, q2 = 0.79, p < .0001) for the five-parameter model. Because the models require only calculated parameters from a 3D chemical structure, they could enable the design or selection of compounds likely to be translaminar.

中文翻译:

杀菌剂跨层活性的修正模型

跨层再分布对于杀菌剂活性很有价值,但难以测量和预测。通过实验测量了 38 种具有抗黄瓜白粉病活性的杀菌剂的跨层活性,并用于根据计算参数开发跨层运动的 QSAR(定量结构-活性关系)模型。考虑了从能量最小化的 3D 结构产生的 300 多个理化参数,并开发了一个参数、两个参数和五参数模型。单参数亲脂性模型解释了完整数据集中 39% 的跨层活性变异性,但没有解释小琥珀酸脱氢酶抑制剂 (SDHI) 组中的任何变异性。将极性表面积参数添加到亲脂性参数将可预测性提高到 52%,并解释了 SDHI 类中超过 70% 的可变性。具有五个理化参数的扩展模型解释了整个跨层再分布中超过 80% 的变异性。三个附加参数与分子大小和反应性相关。这些模型使用留一法验证,该方法对五参数模型具有出色的稳健性(r2adj = 0.83,q2 = 0.79,p < .0001)。由于这些模型只需要从 3D 化学结构中计算出的参数,因此它们可以设计或选择可能是层状的化合物。三个附加参数与分子大小和反应性相关。这些模型使用留一法验证,该方法对五参数模型具有出色的稳健性(r2adj = 0.83,q2 = 0.79,p < .0001)。由于这些模型只需要从 3D 化学结构中计算出的参数,因此它们可以设计或选择可能是层状的化合物。三个附加参数与分子大小和反应性相关。这些模型使用留一法验证,该方法对五参数模型具有出色的稳健性(r2adj = 0.83,q2 = 0.79,p < .0001)。由于这些模型只需要从 3D 化学结构中计算出的参数,因此它们可以设计或选择可能是层状的化合物。
更新日期:2020-05-01
down
wechat
bug