Sanjeevini 3.0: An Enhanced Comprehensive Automated Web Server for Computer-Aided Drug Design

Abstract

Sanjeevini 3.0 represents a significant advancement in the realm of structure-based lead molecule discovery. This enhanced physicochemical principles-based, machine learning-augmented comprehensive software suite/web server version comprises various modules, including RASPD+ and SEARCH-ML (for a rapid virtual screening of large libraries of small molecules against a specified target), AADS and ParDOCK+ (for active site prediction and docking), StackTox (for toxicity prediction of small molecules), and BAPPL+ (for scoring and estimating binding free energies). The pipeline addresses the complex challenges that the global pharmaceutical industry faces in translating a molecular-level understanding of human diseases into a cutting-edge technology for generating suggestions on candidate drug molecules. Each of the modules and the entire pipeline are thoroughly validated on some major lifethreatening diseases, utilizing a dataset of 120 FDA-approved drugs and the corresponding 126 pharmacologically active targets. Remarkably, the entire pipeline requires only 13–15 min to predict candidate drugs for a single target protein. In nearly 90% of the cases, the pipeline successfully rediscovers known FDA-approved drugs for the target proteins. This methodology offers an efficient and streamlined approach without compromising effectiveness. Sanjeevini 3.0 is freely accessible at https://www.scfbio.iitd.ac.in/Sanjeevini/.

Received: 23 December 2025 | Revised: 12 April 2026 | Accepted: 3 June 2026

Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.

Data Availability Statement

The datasets used and/or generated during the current study are available from the corresponding authors upon reasonable request. The PDBbind datasets used for training and validation are publicly available at https://www.pdbbind.org.cn/. The chemical libraries used for screening, including ZINC (https://zinc15.docking.org/), DrugBank (https://go.drugbank.com/), and FDA-approved compounds (https://go.drugbank.com/), are available from their respective public repositories. The StackTox training (ToxBits) and external validation (ToxBits-Val) datasets can be provided by the corresponding authors upon request.

Author Contribution Statement

Dheeraj Kumar Chaurasia: Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Visualization. Aman Sharma: Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Visualization. Madhvi Mishra: Software, Validation, Investigation, Data curation, Writing – original draft, Writing – review & editing. Akanksha Kesharwani: Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft. Raushan Anjum: Software, Validation, Investigation, Data curation. Shashank Shekhar: Software, Resources, Project administration. Aditya Mittal: Software, Resources, Project administration, Funding acquisition. B. Jayaram: Conceptualization, Methodology, Software, Resources, Writing – review & editing, Supervision, Project administration, Funding acquisition.