ISSN print edition: 0366-6352 ISSN electronic edition: 1336-9075 Registr. No.: MK SR 9/7 Published monthly

A hybrid computational framework for antidepressant drug design integrating machine learning algorithms and molecular modeling

Wakeel Ahmed, Tamseela Ashraf, Shahid Zaman, Asad Ullah, and Farhana Khalid

Department of Mathematics, COMSATS University Islamabad, Lahore Campus, Lahore, Pakistan

E-mail: wakeelahmed784@gmail.com

Received: 25 August 2025  Accepted: 26 September 2025

Abstract:

This study presents a novel approach to drug discovery by integrating machine learning algorithms with molecular modeling techniques. Specifically, it focuses on quantitative structure property relationship (QSPR) analysis enhanced by eccentricity-based topological indices derived from molecular structures. The physicochemical properties of antidepressant drugs are predicted using a combination of linear regression, random forest, and XGBoost models. This interdisciplinary framework leverages the strengths of machine learning, computational chemistry, and topological analysis to accelerate and refine the drug design process. The integration of these methodologies not only reduces the time required to identify effective compounds, but also provides deeper insights into drug activity and optimization. The results indicate that this hybrid approach holds significant promise for advancing rational drug design and supporting the development of personalized therapeutic strategies within the pharmaceutical industry.

Keywords: Machine learning algorithms; Antidepressant drugs; Topological indices; Cheminformatics

Full paper is available at www.springerlink.com.

DOI: 10.1007/s11696-025-04413-w

Chemical Papers 80 (1) 589–614 (2026)