Summary:
UCLA researchers in the Department of Microbiology, Immunology, and Molecular Genetics have developed an advanced API gateway that leverages AI to autonomously process data requests with exceptional speed, reliability, and efficiency.
Background:
Application Programming Interface (API) infrastructures are universally employed to enable communication between software applications and services, often requiring transfer of large and sensitive data sets. API utilization supports a range of biomedical research applications, such as ELISA data analysis, medical note-taking, single-cell and flow cytometry analysis, and drug target screening. To ensure data privacy, it is integral for these systems to handle the information effectively with rapid processing and robust security systems. However, existing solutions fall short in these areas due to the lack of efficient mechanisms for intelligent routing, real-time monitoring, security enforcement, and data refinement. These shortcomings highlight the need for a more advanced system—one capable of securely handling sensitive data, dynamically responding to real-time inputs, and delivering faster, more dependable data processing outcomes.
Innovation:
Dr. Lulan Wang has successfully created a novel API gateway system that significantly surpasses previous software in speed, accuracy, and data performance. The system facilitates communication between AI applications and external tools or data sources, allowing for seamless integration with existing research workflows and datasets. In a case study, analysis time per sample was reduced by 90%, from five hours to just 30 minutes. The system demonstrated 100% accuracy in detecting viral genome mutations and accurately mapping them to corresponding protein structures and reference datasets. Additionally, the system utilizes real-time monitoring and AI-driven feedback mechanisms to assess traffic patterns, evaluate system performance, and maintain data integrity—ensuring an optimized and reliable workflow.
This innovation presents a significant advancement in the speed of medical data processing requests, enabling faster, more precise insights for drug development, virus tracking, and biomedical research.
Credit: Lulan Wang
Potential Applications:
● Drug discovery and target screening
● Medical diagnostics and decision making
● Real-time infectious disease and variant analysis
● Viral genome identification
● Protein structure modeling
● ELISA and flow cytometry data interpretation
● Single cell and flow cytometry data analysis
Advantages:
● Improved accuracy in disease and mutation tracking
● Significantly faster results and data analysis
● Automated, AI-driven workflows
● Seamless integration with existing research tools
● Multi-format, comprehensive reporting
● Real-time traffic and system performance tracking
State of Development:
An advanced prototype has been developed; a successful Minimum Viable Product (MVP) has proven high processing speed and analytical accuracy as compared to previous models. Validation of technical feasibility and user experience has been completed on a SARS-CoV-2 analysis workflow.
Incorporation of Third-Party Code:
1. Flask (https://github.com/pallets/flask; BSD-3-Clause; https://github.com/pallets/flask/blob/main/LICENSE.rst)
2. Requests (https://github.com/psf/requests; Apache License 2.0; https://github.com/psf/requests/blob/main/LICENSE), PyMOL (https://github.com/schrodinger/pymol-open-source; BSD-style; https://github.com/schrodinger/pymol-open-source/blob/master/LICENSE)
3. Boto3 (https://github.com/boto/boto3; Apache License 2.0; https://github.com/boto/boto3/blob/develop/LICENSE)
4. OpenAI Python SDK (https://github.com/openai/openai-python; MIT License; https://github.com/openai/openai-python/blob/main/LICENSE)
5. Matplotlib (https://github.com/matplotlib/matplotlib; PSF-based License; https://github.com/matplotlib/matplotlib/blob/main/LICENSE/LICENSE)
6. Zod (https://github.com/colinhacks/zod; MIT License; https://github.com/colinhacks/zod/blob/master/LICENSE)
Integrated Proprietary API Services (not embedded, but dynamically linked):
1. OpenAI GPT API: Proprietary, accessible via https://platform.openai.com.
2. AWS S3 Storage API: Proprietary, accessible via https://aws.amazon.com/s3/.
Open Standards and Protocols:
1. Model Context Protocol (MCP) by Anthropic
Repository URL: https://github.com/anthropics/MCP
License Type: Open Source (MIT License)
License URL: https://github.com/anthropics/MCP/blob/main/LICENSE
Related Papers:
1. https://www.anthropic.com/news/model-context-protocol
2. https://www.infoq.com/news/2024/12/anthropic-model-context-protocol/
3. https://wandb.ai/onlineinference/mcp/reports/The-Model-Context-Protocol-MCP-by-Anthropic-Origins-functionality-and-impact--VmlldzoxMTY5NDI4MQ
Reference:
UCLA Case No. 2025-265
Lead Inventor:
Lulan Wang, Postdoctoral Staff Research Associate, Department of Microbiology, Immunology, and Molecular Genetics