Preliminary prediction of toxicologically relevant physicochemical properties of Novichoks: the first comparative in silico studies

Abstract

Advanced freely available and commercial in silico methods have been applied to predict the key physicochemical properties of Novichok agents, which are supertoxic organophosphorus compounds with limited experimental data due to their extreme toxicity and legal restrictions. Toxicologically relevant physicochemical properties including boiling and melting points, vapour pressure, water solubility, partition coefficients (log P, log K_OW, log K_OA), Henry's constants, and dissociation constants (pKa), were successfully predicted using appropriate in silico methods. These properties are essential for estimating, based on computational models, the toxicological characteristics, potential environmental behavior, and bioaccumulation tendencies of these hazardous substances. The validation process highlighted the strengths and limitations of the applied methods. OPERA and Percepta delivered the most accurate predictions for boiling and melting points, whereas the EPI Suite and TEST excelled in vapour pressure estimates. However, the significant variability in water solubility predictions underscores the need for methodological refinement and consensus approaches. Model predictions suggest that Novichok agents may exhibit high density (1–1.84 g/mL), moderate to high lipophilicity (logP: −0.68 - 3.06), and generally low vapour pressure (log VP: −6.09 - 10 mmHg). Despite these insights, all results should be interpreted as preliminary and hypothesis-generating, pending experimental validation. This study illustrates the potential utility, but also the current limitations, of in silico methods in the context of assessing supertoxic Novichok agents under conditions where experimental work is constrained.