MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202641113185 A) filed by Samreddy Pooja Reddy on September 22, 2026, for Ai-Based Drug Discovery System Using Molecular Property Prediction Models.
Inventor includes Samreddy Pooja Reddy.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: ABSTRACT A computer-implemented system and method for AI-based drug discovery are disclosed. The system comprises a candidate generation module that employs a trained generative machine learning model to propose novel candidate molecular structures, a multi-task molecular property prediction module comprising a trained machine learning model, such as a graph neural network, that predicts a plurality of molecular properties—including binding affinity, physicochemical properties, and toxicity endpoints—together with associated uncertainty estimates, a scoring and ranking module that computes a composite desirability score for each candidate, and an iterative optimization module that applies a closed feedback loop to structurally modify and re-evaluate top-ranked candidates across successive generations. An applicability-domain assessment module flags predictions of reduced reliability for candidates falling outside the training chemical space. The disclosed system reduces the computational and experimental burden of identifying viable drug candidates while improving the accuracy and reliability of molecular property predictions
Disclaimer: Curated by HT Syndication.
