The use of global and local reactivity descriptors of conceptual DFT to describe toxicity of benzoic acid derivatives

• Published in: Computational and theoretical chemistry 2023-08, Vol.1226, p.114211, Article 114211
• Main Authors: Ramírez-Martínez, Cinthia, Zárate-Hernández, Luis A., Camacho-Mendoza, Rosa L., González-Montiel, Simplicio, Meneses-Viveros, Amilcar, Cruz-Borbolla, Julián
• Format: Article
• Language: English
• Subjects: Benzoic acid; cDFT; QTAIM; STR
• ISSN: 2210-271X
• DOI: 10.1016/j.comptc.2023.114211

[Display omitted] •Quantum descriptors correlated benzoic acid-derived compound toxicity via LD50.•STR model descriptors reveal reactivity with enzymes like serine.•STR models with quantum parameters predict toxicity for drug design. The present study explores the use of global and local quantum reactivity descriptors and conceptual density functional theory (cDFT), as well as descriptors of atomic-quantum topology (QTAIM), to describe the structure–toxicity relationship (STR) for 37 compounds derived from benzoic acid, employing the median lethal dose logarithm (log LD50) as a parameter. The method was validated using a coefficient of determination (R2) of 0.8285 and a leave-one-out cross-validation (Q2LOO) of 0.7138, thus fulfilling the statistical criteria indicated for models generated in QSARINS software. The descriptors closely related to the toxicity of the compounds were EHOMO(N−1), ∇2ρ(r)(BCP2), ∇2ρ(r)(BCP5), QH4Hirshfeld, QYY(C1), and ∑q-meta. This finding both enabled the reactivity of the benzoic acid derivatives to be interpreted and facilitated an understanding of the potential chemical bonds with enzymes, such as serine, in a biological environment.