• AI Co-Scientist for Protein Function Prediction and Hypothesis Generation

  Extending the Co-Scientist framework to build reasoning agents for protein function prediction using CEPI and BV-BRC datasets for priority pathogens

• AI-Powered Analysis Pathfinder: From Research Questions to Data and Workflows

  Enabling bench scientists to start with analysis goals and receive both relevant datasets and executable workflows—a complete path from question to execution.

• Assigning Functions to Uncharacterized Genes

  Leveraging machine learning and AI to predict functions of characterized and uncharacterized genes using K-mers, ensemble methods, and LLM embeddings

• Expanding Rhea for Automated Workflow Generation

  Extending the Rhea platform with MCP+RAG to enable automated workflow generation leveraging Galaxy and BV-BRC for infectious disease analysis

• Extracting Influenza Passaging Cell Type from GenBank Records

  Fine-tuning question answering LLMs to extract passaging cell type metadata from unstructured influenza GenBank records

• Extracting Training Data for Automated Workflow Generation

  Creating structured question-answer training datasets from BRC resources and publications to train LLMs for bioinformatics workflow generation

• Generative AI-Driven Workflow Design and Execution via MCP

  Automating bioinformatics workflow design, refinement, and execution using generative AI models guided through the Model Context Protocol (MCP)

• HiPerRAG for Literature-based Data Extraction on Priority Pathogens

  Leveraging high-performance retrieval-augmented generation to extract and curate structured biological data for CEPI priority pathogens from scientific literature

• InterWeb Outbreak Surveillance and Progress Monitoring

  AI-powered outbreak surveillance aggregating web data sources including social media, news, case reports, and public health databases

• PubMed Miner: AI-Powered Sequence Feature Extraction from Literature

  Automated extraction of viral sequence features, mutations, and epitopes from PubMed literature using LLMs to accelerate outbreak response and genomic analysis

• RDF Knowledge Graph Construction

  Transforming Pathogen Data Network Resources to RDF using Large Language Models

• StorySeq: Automated Sequence Narrative Generation

  Automating sequence identification and contextualization using BLAST, database queries, and LLM narrative synthesis to accelerate pathogen and AMR gene discovery

• Viral Structural Phylogenetics

  Optimizing protein language model to predict viral protein structures at scale

• Your Project Here

  Use this template to propose your own codeathon project