Abstract

IN-SILICO STUDIES AIMED AT TOXICITY MODELING OF NITROAROMATICS AND THEIR DERIVATIVES

Tiwari Anupama*

ABSTRACT

Modeled Toxicity of nitroaromatics and their derivatives using three dimensional physicochemical parameters. The proposed set includes 28 molecules of Nitro aromatic compounds. The toxicity of compounds is expressed in terms of LD50 (mg/Kg) for rats and are normalized using -log (LD50). Statistically significant models were derived by MLR using BuildQSAR software. ChemSketch was used to draw the structures. The model obtained by deleting outliers allows us to confirm that the four parametric model is the most significant model with squared correlation coefficient R2 as 0.8477 as far as three-dimensional physicochemical parameters are concerned. If lipophilicity and solubility are considered in addition to the aforementioned parameters, calculated using SwissADMET software then squared correlation coefficient R2 reaches 0.9488. The predictive power was examined using internal and external validation procedures with the help of Minitab software. The value of R2pred is found to be 0.73 and 0.91 respectively, these values imply decent certainty. Conformational analysis was performed through Marvin Sketch 20.9 software with MMFF941 force field for all NAC‟s and it was observed that toxicity values are not showing any scrupulous association with the conformational energy. Nitrobenzene is revealing a well-adjusted combination of all the parameters with excellent scores.[1] Merck Molecular Force Field (MMFF) was developed in the 1990's in the present in-silico studies. 2,4-dinitrophenol and 1,2,3,4,5-pentachloro-6-nitrobenzeneare show best fit for the modified model which has lipophilicity, pharmacokinetic properties and solubility as additions. Proposed prediction is reliable to check any hazard assessment for the fresh set of chemicals.

Keywords: Toxicity Modeling, Nitroaromatics, Physicochemical Parameters, Conformational Analysis, QSAR.