Deep EigenMatics Secures US Patent for AI-Driven Protein Design

Deep EigenMatics has been awarded a United States Patent for its groundbreaking AI method in protein design and drug discovery, announced in a press release. The Texas-based startup, founded in February 2025, has developed a two-headed neural network that simultaneously generates the sequence and structure of proteins, a significant advancement over previous methods that primarily focused on predicting structure from sequence.

The patented method, invented by Dr. Stephen G. Odaibo, aims to enhance the efficiency of drug discovery by addressing the dual characteristics of proteins—sequence and structure—that determine their function. This innovation is expected to generate promising candidate ligands for drug development across various therapeutic areas, including cardiovascular, oncology, immunology, neurology, and ophthalmology.

Deep EigenMatics is actively engaging with large pharmaceutical companies for potential licensing and co-development partnerships, while continuing to expand its intellectual property portfolio globally. The company is also utilizing its proprietary platform, Deep Discovery™, to initiate the generation of candidate peptide and small molecule drug ligands.