Abstract

MECHANISTIC TRENDS IN ANTICANCER ACTIVITY OF SOME KNOWN DIHYDROPHENANTHRIDINETRIONES (DPT); A CLUSTERED QSAR ANALYSIS

Mohamed Osman M. A. El-Fakii*, Aia Alfatih Musa and Rajaa Satti Mohammed

ABSTRACT

The anticancer activity trends of 24 known 3,4-dihydrophenanthridine-1,7,10(2H)-triones were analyzed through QSAR methodology. Their anticancer activity against human gastric adenocarcinoma (AGS) cell line as IC50 was regressed with the logarithm of their octanol/water partition coefficient (log P) and molar refractivity (MR). Statistical metrics show poor correlation (R2=0.14 for Log P and R2=0.37 for MR and R2 =0.38 for both Log P and MR combined). As these compounds show good activity against AGS cell line, clustering was utilized to expose possible mechanistically distinct subgroups nesting inside the main data set. This was done through a kind of a previously reported regression clustering procedure based on log P which was chosen owing to its wide-spread use in QSAR. This procedure was found to yielded four clusters with high R2 values (~ 0.8) which exhibit distinct trends in their respective regression lines suggesting different operational mechanism for each. Molar refractivity was used afterwards to give depth to the clustering procedure through emphasizing the distinction of each clusters from the others.

Keywords: dihydrophenanthridinetriones (DPT), anticancer activity, clustered regression analysis, QSAR.