INSILICO MODELLING ON SOME C14-UREA TETRANDRINE COMPOUNDS AS POTENT ANTI-CANCER AGAINST HUMAN ERYTHROLEUKEMIA (HEL) CELL LINE

Abstract

Insulin modeling was performed on 28 C14-urea tetrandrine compounds as inhibitors of leukemic (HEL) cell lines using Quantitative Structure-Activity Relationship (QSAR) method. The structure of the inhibitors was correctly drawn, then geometrically optimized at Density Functional Theory (DFT) level (DFT / B3LYP / 6-31G *) with Spartan 14 V1.1.4. Also, molecular descriptors of the inhibitors were calculated with PaDEL calculator, and the results were partitioned into training and test set after data pretreatment. The training set was used to generate a model by employing genetic function approximation in choosing best descriptors to form the model. The validation parameters of the model include; R ^ 2 (train) at 0.8067, LOF 0.037 r ^ 2 (QCV) to 0.6378 R ^ 2 (test) 0.7629 of the CRP ^ 2 and the 0. 6990 Which have passed the acceptance criteria for a QSAR model worldwide. In addition, the model depicted four (4) descriptors, AATS4v, AATS5i, AATSC5i, and GATS5m with positive meanings signifying that increase in these descriptors will positively influence and increase the activity of the inhibitors. This study depicts a route in designing and synthesizing new C14-urea tetrandrine compounds with better inhibitory potentials.

Keywords

• QSAR
• Mean Effect
• Validation
• Descriptors
• Model
• Y-randomization