Exploring 3-arylquinoxaline-2-carbonitrile 1,4-di-N-oxides activities against neglected diseases with QSAR.

This work reveals our efforts to continue identifying new active compounds against neglected diseases, such as malaria and tuberculosis. We took several 3-arylquinoxaline-2-carbonitrile 1,4-di-N-oxide derivative activity results from the reference literature and established useful quantitative structure-activity relationships. We hoped that the development of in silico models would broaden our knowledge regarding the overwhelming problem of drug resistance to both illnesses. The optimized molecular structures of 60 compounds were represented by 1497 DRAGON-type descriptors and subjected to linear regression analyses; the quantitative structure-activity relationships resulted predictive when searching for new active compounds. We obtained a rational guide for the proposal of new candidate structures whose activities still remain unknown.

QSAR analysis on Spodoptera litura antifeedant activities for flavone derivatives.

We establish useful models that relate experimentally measured biological activities of compounds to their molecular structure. The pED(50) feeding inhibition on Spodoptera litura species exhibited by aurones, chromones, 3-coumarones and flavones is analyzed in this work through the hypothesis encompassed in the Quantitative Structure-Activity Relationships (QSAR) Theory. This constitutes a first necessary computationally based step during the design of more bio-friendly repellents that could lead to insights for improving the insecticidal activities of the investigated compounds. After optimizing the molecular structure of each furane and pyrane benzoderivative with the semiempirical molecular orbitals method PM3, more than a thousand of constitutional, topological, geometrical and electronic descriptors are calculated and multiparametric linear regression models are established on the antifeedant potencies. The feature selection method employed in this study is the Replacement Method, which has proven to be successful in previous analyzes. We establish the QSAR both for the complete molecular set of compounds and also for each chemical class, so that acceptably describing the variation of the inhibitory activities from the knowledge of their structure and thus achieving useful predictive results. The main interest of developing trustful QSAR models is that these enable the prediction of compounds having no experimentally measured activities for any reason. Therefore, the structure-activity relationships are further employed for investigating the antifeedant activity on previously synthesized 2-,7-substituted benzopyranes, which do not pose any measured values on the biological expression. One of them, 2-(alpha-naphtyl)-4H-1-benzopyran-4-one, results in a promising structure to be experimentally analyzed as it has predicted pED(50)=1.162.