MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202641119005 A) filed by Sreyas Institute Of Engineering And Technology on October 05, 2026, for Federated Learning-Driven Decentralized Ai System For Privacy-Preserving Collaborative Drug Discovery.
Inventors include Dr. A Pradeep; A Kanakaiah; and D Rajashekar Reddy.
The application for the patent was published on October 09, 2026, under issue no. 41/2026.
Abstract: ABSTRACT [0014] A federated learning-driven decentralized artificial intelligence system is disclosed for privacy-preserving collaborative drug discovery. Multiple independent research entities maintain local private datasets of molecular structures, biological assay results, and clinical observations. Each entity trains a local neural network model on its own data without transferring raw records. Only encrypted model parameter updates are exchanged through a secure aggregation protocol coordinated by a blockchain-anchored orchestrator. The orchestrator computes a global model that improves predictive accuracy for drug-target interactions, toxicity profiles, and candidate ranking while ensuring that no participant can reconstruct another participant’s data. The system incorporates differential privacy noise injection, homomorphic encryption of gradients, and adaptive client selection based on data quality and contribution history. Experimental validation on simulated multi- institutional cohorts demonstrates accelerated identification of novel molecular candidates with substantially reduced risk of privacy leakage compared with centralized training approaches.
Disclaimer: Curated by HT Syndication.
