A study based on binary fingerprints for flavone analogues as CDK2/Cyclin a inhibitors. a two dimensional QSAR approach

In the present study, we have applied seven available similarity based binary fingerprints as implemented in the new cheminformatics package Canvas, on a validated dataset for the test compounds selectively inhibiting CDK2/Cyclin A. The fingerprint methods used were: Linear, Dendritic, Radial, MOLPRINT2D, Pairwise, Triplet, and Torsion. Out of the seven fingerprints used, the fingerprint dentritic resulted in a statistically significant 2D QSAR model with regression coefficient (r2) value of 0.9284 and cross validation coefficient (q2) value of 0.9865. The model could be used to design potent inhibitors against the target CDK2/Cyclin A as a goal towards development of novel anticancer agents.