MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202641112415 A) filed by Sravathi Ai Technology Private Limited on September 18, 2026, for Non-Monotonic-Piecewise-Linear Mapper For Predicting Protein-Ligand Binding Affinity For In-Silico- Drug-Discovery.
Inventors include Ben Geoffrey A S; and Abhishek Singh.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: The instant invention discloses Prediction of protein-ligand binding affinity employing fundamental physical principles such as conservation of energy during optimization. Physics-informed neural networks (PINNs) have been proposed aiming at embedding physical laws directly into learning architectures and training protocols. The developed approach introduces a novel architecture and training scheme wherein an equivalence is established between solving for the thermodynamic action integral and minimizing the loss function of the problem. In the developed scheme, novel trainable components such as piece-wise linear functional mappers that can learn an unknown functional correlation between input features and target variable of interest are disclosed. More particularly, the invention discloses computer- implemented systems and methods for predicting protein-ligand binding affinity using physics-informed machine learning models that integrate molecular mechanics- derived physical descriptors with trainable functional mapping architectures for improved prediction accuracy and generalization across chemically diverse compounds especially in in-silico drug discovery.
Disclaimer: Curated by HT Syndication.
