GeneInsight: Condensing gene set knowledge via language models.

Gene set analysis often returns extensive annotations from multiple sources, requiring manual effort to identify coherent biological themes. We developed GeneInsight, an AI-powered tool that automates this by retrieving functional annotations from STRING-DB, clustering semantically related terms using sentence embeddings, and generating thematic summaries via large language model prompting. This enables researchers to identify biological themes that may be obscured when annotation sources are examined separately.