Abstract

A REVIEW ARTICLE ON QSAR APPLICATION AND ADVANCEMENT IN DRUG DESIGN

Miss. Prerna S. Deodhagale*, Prof. Suraj P. Rajurkar, Miss. Shruti P. Kadam, Mr. Priyanshu Y. Chaudhari, Miss. Sakshi B. Deshmukh, Mr. Nikhil R. Rathod

ABSTRACT

Quantitative Structure-Activity Relationship (QSAR) modeling is a fundamental approach computational chemistry that uses mathematical and statistical methods to connect the structural characteristics of chemical compounds to their biological activities or physicochemical properties. By converting molecular structures into numerical descriptors, QSAR allows researchers to predict the activity of untested compounds, reducing the need for extensive experimental screening. Over the last few decades, QSAR has progressed from simple linear models to complex machine learning and deep learning frameworks, significantly improving predictive performance and applicability across chemical and biological systems. This review offers a thorough analysis of QSAR methodologies, including contemporary artificial intelligence-based algorithms, nonlinear modeling approaches, and traditional linear regression techniques. The descriptor calculation, feature selection techniques, model validation, and applicabilitydomain assessment—all crucial elements that affect the generalizability and dependability of QSAR models—are given particular attention. Recent developments in fragment-based QSAR, 3D-QSAR, and hybrid modeling techniques are also covered, emphasizing how they can speed up materials design, toxicity assessment, and drug discovery.

Keywords: Quantitative structure activity relationship (QSAR), computational chemistry, QSAR application, 3D-QSAR, machine learning in QSAR, structure activity relationship (SAR).