Interactive Systems and Methods for Identifying Target Proteins in Drug Discovery (Case No. 2025-098)

Summary:

UCLA researchers from the Department of Electrical and Computer Engineering have developed a novel computational system for target protein identification, enabling integrative drug discovery.

Background:

Target identification (Target ID) in drug discovery involves the identification and evaluation of protein candidates that could interact with disease-associated proteins and therapeutically relevant ligands. This complex process is computationally intensive and requires the evaluation of protein-protein interaction networks and their role in disease development. This process also involves predicting and assessing the affinity between proteins and associated ligands. ALS, Alzheimer's disease, and Parkinson's disease are examples of disorders linked to protein misfolding, where researchers are working to identify therapeutically relevant proteins for potential disease mitigation. Existing methods of Target ID rely on several unintegrated tools to assess criteria such as interaction networks, therapeutic relevance, and binding affinity. These methods function independently, therefore requiring manual data integration and repeated system switching, both of which introduce human error and cognitive strain, and limit scalability of these computational techniques. There is a continuing need for a unified Target ID tool that streamlines this computational process, enabling researchers to more effectively scale target protein discovery with potential therapeutic applications.

Innovation:

UCLA researchers from the Department of Electrical and Computer Engineering have developed a modular computational tool that unifies the Target ID process within a single interaction platform. The reported technology, titled HAPPIER: (Human-AI Protein-Protein Interaction Discovery), integrates multiple computational models within one platform, including semantic similarity modeling, retrieval-augmented generation with large language models, and docking simulations. This innovation allows for the prediction of the therapeutic impacts of different proteins and the spatial affinities between these protein targets and various ligands. The modular design of this tool allows users to generate multi-layered protein interaction graphs, assess therapeutic potential using biomedical databases, and simulate ligand-protein binding across multiple proteins. This high throughput method further allows users to rank potential protein targets based on user-defined weights assigned to different criteria including docking score and therapeutic impact. HAPPIER has the potential to revolutionize drug discovery by integrating disparate computational models into a modular system that allows for the identification and evaluation of therapeutically relevant protein targets.

Potential Applications:

  • Target identification for neurological diseases
  • Oncology drug discovery
  • Structure-based drug design
  • High-throughput screening

Advantages:

  • End-to-end integration of AI models
  • Interactive multi-criteria prioritization
  • High throughout simulations
  • Modular and scalable
  • Adaptable to therapeutic fields
  • Reduced cognitive load

Development-To-Date:

First successful description of the invention (06/01/2024)

Reference:

UCLA Case No. 2025-098

Lead Inventor:

Xiang Anthony Chen

Patent Information:
For More Information:
Joel Kehle
Business Development Officer
joel.kehle@tdg.ucla.edu
Inventors:
Xiang Chen
Youngseung Jeon
Christopher Hwang
Ziwen Li
Jesus Campagna
Varghese John
Whitaker Cohn
Eunice Jun