Classification, QSAR analysis and molecular docking of some α-naphthoflavone derivatives as potent CYP1B1 inhibitors

Cytochrome P450 (CYP) enzymes are a group of hemoprotein monooxygenases that participate in the biotransformation of various substances. These enzymes can naturally produce in the body or by external chemicals, such as commonly prescribed medications like docetaxel, tamoxifen, and cisplatin [1]. In the present study, classification and quantitative structure-activity relationship (QSAR) models has been employed to predict inhibitory concentration values and determine active/inactive inhibitors. The models were developed utilizing a dataset consisting of 41 inhibitors of the CYP1B1 enzyme based on α-naphthoflavone derivatives [2]. The dataset was split into the calibration and prediction sets to develop and validate the obtained models, respectively. The active/inactive inhibitors were classified using partial least squares-discriminant analysis (PLS-DA) and k-nearest neighbors (kNN) techniques and molecular descriptors [3]. The specificity, sensitivity and precision of PLS-DA technique are all more than 90% for classification of CYP1B1 inhibitors. The QSAR analysis was performed using stepwise multiple linear regression (stepwise MLR), partial least square (PLS) and support vector machine (SVM). The results of QSAR analysis shows satisfactory results in terms of adjusted R2 ( 0.85) and standard error ( 0.5) for all three stepwise MLR, PLS and SVM techniques [4]. Also, Molecular docking studies were used to identify both the important binding sites on the CYP1B1 enzymes and crucial hetero atoms of inhibitors. By utilizing these computational approaches, the current study aimed to assess inhibitory concentrations, interactions and binding patterns of these inhibitors with the CYP1B1 enzyme, eventually providing both predictive and descriptive QSAR analysis.