MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202641124927 A) filed by Pes University on October 19, 2026, for System And Method For Biopharmaceutics Classification Using Molecular Features And Predicted Aqueous Solubility.
Inventors include Hari Kiran K; Hari Shankar; Krupa S; Rhushya K C; Prajwala Tr; and Saniya Jawed.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: The present disclosure provides a system (100) and method (200) for predicting a Biopharmaceutics Classification System (BCS) class from molecular structure. A processor (104-A) executes instructions stored in a memory (104-B) to validate a received SMILES string using an input validation unit (106). A descriptor extraction unit (108) computes molecular descriptors including LogP, and a fingerprint generation unit (110) generates an ECFP4 fingerprint vector. A solubility prediction unit (112) employs a trained regression model (112-A) to obtain a predicted LogS value. A feature assembly unit (114) combines the predicted LogS value, LogP and ECFP4 representations into a combined feature vector. A BCS classification unit (116) supplies the combined feature vector to a trained ensemble classifier (116-A) to predict a BCS class. The models are trained independently and operate sequentially during inference. FIG. 1 illustrates the system (100).
Disclaimer: Curated by HT Syndication.
