Abstract

FOLLOW UP OF THE CYTOTOXICITY OF SOME DIHYDROPHENANTHRIDINE-1, 7, 10(4H)-TRIONES AGAINST SK-MES-1 HUMAN LUNG CANCER CELL LINE VIA QSAR

Maeen Ahmed M. A. AL-Mansori*, Mohamed Osman M. A. El-Fakii and Inas O. Khojali

ABSTRACT

In support of previously reported clustering methodology used to unearth possible mechanistically distinct subgroup within a data set of known dihydrophenanthridinetriones (DPT), we chose a number of descriptors to be superimposed into MR-based clustering. This is done to emphasize the distinction among different clusters through regression. The chosen descriptor which include quantum mechanically computed electrophilicity index (ω), heat of formation (Hf), enthalpy difference of quinone-hydroquinone forms of DPT (ΔH) together with its reciprocal (1/ΔH). We also choose to include one topological descriptor (Balban index J) and one classical descriptor (total of atomic polarizabilities apol). The depth of impact of each descriptor on activity was assessed by regression analysis and contemplating the value of Pearson coefficient. The mutual dependence of descriptors for each cluster was also studied via constructing correlation matrix using the correlation coefficient as indicator of property space overlapping; thus, a unique profiling of each cluster was arrived at boosting our previous clustering outcome for DPTs.

Keywords: DPT, QSAR, clustering, mechanistic profile.