MUMBAI, India, Jan. 23 -- Intellectual Property India has published a patent application (202531087308 A) filed by Jis University, Kolkata, West Bengal, on Sept. 14, 2025, for 'ai-driven ensemble platform for accurate and interpretable drug activity prediction using shap-enhanced model stacking.'
Inventor(s) include Mr. Amit Kumar Bhore; Dr. Ananjan Maiti; and Dr. M. Nazma Naskar.
The application for the patent was published on Jan. 23, under issue no. 04/2026.
According to the abstract released by the Intellectual Property India: "The disclosure discloses an artificial intelligence-driven, ensemble-based drug activity prediction platform configured to perform automated, reproducible, and interpretable virtual screening of chemical compound libraries. The system integrates multiple heterogeneous base learners, including Random Forest, Support Vector Machine, and XGBoost, into a stacked ensemble architecture wherein a logistic regression meta-learner synthesizes probabilistic outputs to achieve superior predictive accuracy and robustness. Automated preprocessing, normalization, and imputation modules ensure chemical diversity is preserved. Integrated SHAP analysis provides global and local interpretability by quantifying the contribution of molecular descriptors to predictions. The end-to-end pipeline archives all metrics, visualizations, and intermediate results, thereby enabling transparency, regulatory compliance, and scalability. The disclosure thus addresses deficiencies of prior art by harmonizing predictive fidelity, interpretability, and automation, thereby de-risking early-stage drug discovery and significantly reducing experimental costs."
Disclaimer: Curated by HT Syndication.
