Quantitative Structure Activity Relationship Studies of Potent Endothelin-A Receptor Antagonist for the Treatment of Pulmonary Arterial Hypertension

Abstract

A traditional physicochemical descriptors-based QSAR analysis was conducted on a data-set of Endothelin-A Receptor Antagonists for the treatment of Pulmonary Arterial Hypertension. A variety of statistical techniques, including non-linear techniques like artificial neural networks and linear analytical techniques i.e., ‘Multiple Linear Regression’ and ‘Partial Least Squares’ implicated in current research. The development models were then put through a validation process including the leave one out, which supported their high predictability and accuracy. A few statistical parameters were used to build the model’s predictive power and the resulting model was found to have good statistical values, such as s=0.40, f=48.75, r=0.87, r²=0.77, r²CV=0.71 for training set. Three descriptors, including logP (whole molecule), total lipole(whole molecule), VAMP LUMO(whole molecule)were made relevant by the general model, which offered insightful information. As a new results, these traits may be successfully used to the modelling and screening of new endothelin-A receptor antagonists that are active hypertensive drugs.

Keywords: Endothelin-A receptor antagonists, Quantitative- structure activity relationship, statistical analysis, pulmonary arterial hypertension