Licensed under a Creative Commons Attribution 4.0 International (CC BY 4.0)

The identification of molecular descriptors that embody the chemical information for druglikeness will be a step forward in data-driven drug discovery and development endeavor. In this study, over 4000 Dragon-type molecular properties were generated for approximately 2000 known drugs and 2000 surrogate nondrugs. Logistic Regression (LogR) and Random Forest (RF) techniques were carried out to unveil the crucial molecular descriptors that can adequately classify a compound as drug or nondrug. Ten one-variable LogR models each demonstrated at least 70% prediction accuracy. A two-variable model consisting of HVcpx and MDDD correctly classified 85% of the test compounds. The best LogR model with 89.0% prediction accuracy identified five most influential descriptors for druglikeness: an information index HVcpx, topological index MDDD, a ring descriptor NNRS, X2A or average connectivity index of order 2, and walk and path count SRW05. The best RF model involving 10 only weakly correlated descriptors was found to be 92.5% accurate and at par with the RF and LogR models that consisted of over 200 variables. The model featured: molecular weight, MW; average molecular weight, AMW; rotatable bond fraction, RBF; percentage carbon, C%; maximal electrotopological negative variation, MAXDN; all-path Wiener index, Wap; structural information content index, neighborhood symmetry of 1 order, SIC1; number of nitrogen atoms, nN; 2D Petitjean shape index, PJI2; and self-returning walk count of order 5, SRW05. Many of these descriptors have straightforward chemical interpretability and future applicability as druglikeness filters in virtual high throughput drug discovery.

The discovery of next-generation non-steroidal anti-inflammatory drugs (NSAIDs) remains an active area of research as over a billion people suffer from pain and inflammation. A strategic approach in this endeavor is establishing a quantitative relationship between the anti-inflammatory activity and the molecular descriptors of inhibitors of cyclooxygenase-2 (COX-2) that will streamline and expedite the discovery and the subsequent development of novel NSAIDs devoid of side effects associated with COX-1 inhibition. In this work, Random Forest (RF) technique was implemented to formulate a robust quantitative model that predicts the inhibitory activity of compounds on COX-2. The model established in this work displayed excellent predictive performance on compound classification with 93% accuracy and 0.98 AUC. Upon application to two external sets, 759 newly designed derivatives of COX-2 inhibitors and 188 structurally similar compounds were predicted active; 19 of them were found to be promising leads as COX-2-acting anti-inflammatory drugs. The top 2 hits with the highest probability of being active were also found to have the strongest binding affinity with COX-2 and are superior to the known COX-2 selective inhibitors. The RF model is likewise conservative in identifying compounds as active making it all the more beneficial as it helps avoid costly failures at the later stages of the drug discovery phase.This is an open-access article distributed under the terms of the Creative Commons Attribution-Non Commercial-Share Alike 4.0 License, which allows others to remix, tweak, and build upon the work non commercially, as long as the author is credited and the new creations are licensed under the identical terms.

Gout and oxidative stress have been strongly associated with hyperuricemia, a metabolic defect marked by high levels of uric acid (UA) in the serum. Hyperuricemia has been managed by the use of drugs that inhibit xanthine oxidase. The recent account on synthesis of 4-aryl/heteroaryl-4H-fused pyrans as XO inhibitors provided excellent opportunity to uncover the crucial properties of these compounds that confer XO inhibitory action. In here, multiple linear regression analysis of DRAGON-type descriptors showed that the Randic Shape Index (PW3), and a size descriptor P_VSA_v_3account for the 75% of the variability of IC 50 values. Correlation studies with familiar QSAR descriptors indicate that the observed activity is primarily influenced by the molecular ovality and volume, and partly by charge distribution. The Comparative Molecular Field Analysis (CoMFA) models provide further insights on the steric and electronic features of this class of XO inhibitors.

No abstract