Multi-Platform In Silico Assessment of Selected A-Series (Novichok) Nerve Agents: Physicochemical, Toxicological, and Mechanistic Insights

A multi-platform in silico framework was developed to characterize five selected A-series (Novichok) nerve agents—A-230, A-232, and A-234 (each evaluated using Ellison- and Mirzayanov-type structural representations) [1,2] together with A-242 and A-262—through integrated physicochemical, toxicological, and biological-activity prediction. Four complementary computational platforms (ACD/Labs Percepta, OECD QSAR Toolbox, EPA T.E.S.T., and PASS Online) were combined within a reproducible workflow to support Structure–Property–Toxicity Relationship (SPTR) interpretation across eight modeled structures (n = 8). Predicted partitioning behavior indicated moderate lipophilicity (logP (GALAS) ≈ 0.5–1.7) together with very low vapor pressure (0–0.05 mmHg at 25 °C), consistent with condensed-phase persistence and contact-driven exposure rather than sustained vapor dispersion. QSAR-derived LD₅₀ values and dermal permeability coefficients varied across structural variants, demonstrating that predicted acute toxicity is not governed by a single descriptor. PASS-derived probabilities indicated elevated neurotoxicity likelihoods, while cholinergic activity probabilities showed stronger association with polarity-related descriptors (TPSA) than with lipophilicity alone. Principal Component Analysis (PC1 ≈ 81%, PC2 ≈ 13%) showed that descriptor variance was primarily driven by polarity–permeability balance rather than toxicity magnitude alone. Composite hazard ranking integrating inverse LD₅₀, permeability, and neurotoxicity probability identified structural variants combining higher predicted permeability and elevated activity probabilities as comparatively higher within the modeled set. All findings are restricted to the studied compounds and represent computational predictions rather than experimental measurements, providing a transparent and regulatory-aligned framework for comparative, non-experimental hazard assessmen under Chemical Weapons Convention constraints. • A multi-platform in silico framework was applied to five A-series nerve agents. • Physicochemical and toxicological endpoints were cross-validated across four tools. • A-series agents show intermediate volatility and V-series–like environmental persistence. • Strong structure–property–toxicity relationships were identified (r > 0.85, p < 0.05). • Results support regulatory-aligned, non-experimental hazard assessment of CWAs.