Tevogen Bio Holdings Inc. (NASDAQ: TVGN) announced that the World Intellectual Property Organization has published its international patent application for artificial intelligence technology that predicts immunologically active peptides.

The patent application, titled "Systems and Methods for Predicting Immunologically Active Peptides with Machine Learning Models," covers technology developed by Tevogen.AI that uses machine learning algorithms powered by Microsoft Corp. (NASDAQ: MSFT) and Databricks to identify peptides with immune system interactions.

The technology aims to screen and rank potential peptides based on their immunological activity, eliminate peptides likely to be ineffective due to self-tolerance or human genome overlap, and refine predictions using real-world data to train machine learning models.

The Warren, New Jersey-based company stated that traditional methods for identifying immunologically active peptides face limitations, including overlooking human genetic diversity factors such as age, sex, race, and ethnicity.

"Leveraging AI to accelerate discovery, shorten development timelines, and reduce costs is essential to our mission of delivering commercially attractive, economically viable, and cost-effective personalized T cell therapies," said Dr. Ryan Saadi, founder and CEO of Tevogen.

The patent publication carries the designation WO 2025/129197. Identifying immunologically active peptides is considered important for developing targeted therapies for diseases including cancers and infectious diseases.

The information was disclosed in a company press release statement.