Supplementary material of "Anti-inflammatory boosted QSAR models for predicting Nitric Oxide inhibition by bioactive metabolites from Aspergillus species""

The study focuses on employing advanced QSAR modeling techniques to predict NO inhibitory metabolites from Aspergillus. It utilizes decision trees, the Ranker method, and CorrelationAttributeEval for attribute selection, enhanced by Rotation Forest and AdaBoost, with classifiers like Artificial Neural Networks and J48 Trees. The QSAR models, using an ensemble approach with Rotation Forest and AdaBoost.M1 in a KNIME workflow, selected seven molecular descriptors from a dataset of diverse Aspergillus anti-inflammatory metabolites. The RF-enhanced model excelled in capturing complex relationships and enhancing predictive performance. It identified electrotopological state, topological distances, and secondary amides as key contributors to NO inhibition. This model shows promising potential for identifying new anti-inflammatory drugs based on computational techniques.