Quantitative structure activity relationship for the computational prediction of nitrocompounds carcinogenicity

Area of Science:

• Computational toxicology
• Medicinal chemistry
• Structure-activity relationship analysis

Background:

• Rodent carcinogenicity bioassays are lengthy, expensive, and ethically concerning due to animal usage.
• Quantitative structure-activity relationships (QSARs) offer a promising alternative for toxicity prediction and reducing animal testing.
• There is a need for reliable and transparent methods to predict chemical carcinogenicity.

Purpose of the Study:

• To develop and validate a QSAR model for predicting the carcinogenicity of nitrocompounds in rodents.
• To assess the predictive performance of the TOPological substructural molecular design (TOPS-MODE) approach for nitrocompound carcinogenicity.
• To elucidate the structural determinants of carcinogenicity in nitrocompounds.

Main Methods:

• Application of the TOPS-MODE (TOPological substructural molecular design) approach to a dataset of nitrocompounds.
• Model development and validation using leave-one-out cross-validation.
• Comparative analysis of the TOPS-MODE model against four other predictive models.

Main Results:

• The TOPS-MODE model explained 79.10% of the experimental variance in rodent carcinogenicity (standard deviation = 0.424).
• Leave-one-out validation yielded a determination coefficient of 0.666, confirming the model's predictive power.
• The model identified specific structural features, such as primary amine groups and the type of carbon atom bonded to the nitro group, influencing carcinogenic potency.

Conclusions:

• The TOPS-MODE approach provides a transparent and effective method for predicting nitrocompound carcinogenicity.
• This QSAR model significantly reduces the need for animal testing in carcinogenicity assessments.
• The identified structure-activity relationships offer valuable insights into the mechanisms of nitrocompound-induced carcinogenicity.