Generating Understanding and Interpretation of Multi-omics Data with an Automated and Generalizable Pipeline (GENRAItOR)
PI: Sam Erwin, National Security Directorate
GenrAItOR seeks to use GenAI for identification of host biological mechanisms from key features extracted during multi-omics data harmonization. This is accomplished using RAFT fine-tuning to construct an LLM that will accept queries about the relationship between key features and output descriptions of potential relationships identified in the omics literature.