Rough Set Theory as an Interpretable Method for Predicting the Inhibition of Cytochrome P450 1A2 and 2D6

Abstract

Early prediction of ADME properties such as the cytochrome P450 (CYP) mediated drug-drug interactions is an important challenge in the drug discovery area. In this study, we propose to couple an original data mining approach based on Rough Set Theory (RST) to a structural description of molecules. The latter was achieved by using two types of structural keys: (1) the MACCS keys and (2) a set of five in-house fingerprints based on properties of the electron density distributions of chemical groups. The compounds considered are involved in the inhibition of CYP1A2 and CYP2D6. RST allowed the extraction of rules further used as classifiers to predict the inhibitory profile of an independent set of molecules. The results reached prediction accuracies of 90.6 and 88.2 % for CYP1A2 and CYP2D6, respectively. In addition, these classifiers were analyzed to determine which structural fragments were most used for building the rules, revealing relationships between the occurrence of particular molecular fragments and CYP inhibition. The results assessed RST as a suitable tool to build strongly predictive models and infer structure-activity rules associated with potency.

Perspectives on Knowledge Discovery Algorithms Recently Introduced in Chemoinformatics: Rough Set Theory, Association Rule Mining, Emerging Patterns, and Formal Concept Analysis

“…Table 2 shows details of some substructures contained within the molecules, together with a classification of the molecules as active or not active. Table 2 is a decision table where U is the set of molecules, U = {1,2,3,4,5,6,7,8} and C is the set of conditional attributes which comprise the 18 It is important to discover the degree of dependence between attributes 19 and the degree to which the decision class, D, depends upon the conditional attributes, C. D depends on C in a degree k (0 ≤ k ≤ 1), where k represents the fraction of objects which can be correctly classified in D using the conditional attributes in C. Formally,…”

“…Their presence could, for example, mean experimental error leading to misclassification or it could indicate the presence of an activity cliff. 18 It is important to discover the degree of dependence between attributes 19 and the degree to which the decision class, D, depends upon the conditional attributes, C. D depends on C in a degree k (0 ≤ k ≤ 1), where k represents the fraction of objects which can be correctly classified in D using the conditional attributes in C. Formally,…”

“…Very recently, RST has been reported as a method for the prediction of Cytochrome P450 inhibition. 18 The main attraction of the method is the interpretability of the results. Compounds were represented both as MACCS keys and as inhouse electron density-derived fingerprints, and RST was used to generate rules both to classify compounds and to predict the inhibitory profile of unknown molecules.…”

“…The presence of molecules in the boundary region can suggest the need for further investigation. Their presence could, for example, mean experimental error leading to misclassification or it could indicate the presence of an activity cliff …”

“…Very recently, RST has been reported as a method for the prediction of Cytochrome P450 inhibition . The main attraction of the method is the interpretability of the results.…”