Drug-likeness, herbicide-likeness and toxicity of herbicidal compounds – in silico analysis
Vesna Pehar, Davor Oršolić, Višnja Stepanić · 2019
Instead of conventional process of discovery of novel drugs which is complex, time-consuming and expensive, revealing novel targets and therapeutic applications of existing drugs is an alternative and efficient approach used for increasing productivity in the pharmaceutical industry. We explored the data set of 509 approved herbicides for their use as a chemical library for finding novel lead molecules. Herbicides were analysed in terms of drug-likeness by applying common in silico filters for nondrug-like molecules. Descriptors were calculated by the program DataWarrior and analysis was performed by using R packages and Excel. The herbicides were compared with 2106 drugs. Most of herbicides are drug-like compounds and the set of structurally diverse herbicides is recommended as a screening library for medicinal chemistry projects. Herbicides are neutral or negatively charged molecules. Their clogP and TPSA values are mainly in the ranges 0.5-3.5 and (20-100) A2, respectively, which are possessed by around 1/3 of drugs. However, the successful application of herbicides in medicine may depend upon their toxicity for humans. More than 2/5 of the herbicides are estimated to be mutagenic, carcinogenic, toxic for reproduction and irritant. In comparison, around 1/5 of drugs are estimated to be toxic.