I-Corps: Translation Potential of an Extensible Approach for Microbial Profiling from Sequencing Data

Award ID: 2507498

The significance of this I-Corps project is based on the translation from lab to market of an artificial intelligence (AI)-powered microbial profiling platform that analyzes DNA sequencing data with enhanced speed, accuracy, and scalability. The benefits of this approach include increased scalability, reduced processing time, and improved sensitivity for detecting rare microbial species. Microbial profiling is crucial for applications in public health, infectious disease and environmental monitoring, precision medicine, and pharmaceutical development. For example, in the U.S. alone, healthcare-associated infections affect over 1.7 million patients annually and current sequence analysis workflows are too resource-intensive to scale effectively, creating a significant barrier to timely and cost-efficient microbial detection. By reducing the computational burden and expertise required for sequencing analysis, this innovation may expand access to genomic insights, facilitating advances in healthcare, biotechnology, and biosecurity while reducing costs for clinical and research institutions. This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of a microbial profiling platform powered by large language models (LLMs). This approach processes DNA sequencing data as a structured language, applying deep learning techniques to recognize taxonomic patterns, host-microbe interactions, plasmid differentiation, and antibiotic resistance markers. This new solution consists of a two-step artificial intelligence (AI) framework: (1) foundational LLMs pre-trained on extensive sequencing datasets to learn genomic patterns and (2) a fine-tuning mechanism that customizes model predictions for domain-specific microbial profiling. Unlike conventional alignment-based sequencing analysis which requires extensive computational resources and bioinformatics expertise, this method rapidly extracts meaningful biological insights with minimal manual intervention and further automates data interpretation, optimizing results for clinical and research applications. The ability to rapidly analyze microbial communities with high accuracy has significant implications for infectious disease management, antimicrobial resistance tracking, and precision medicine applications. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Program Director: Ruth Shuman