TeroSeek: An AI-Powered Knowledge Base and Retrieval Generation Platform for Terpenoid Research

Terpenoids repre sent a pivotal class of natural products that have garnered su stained scientific interest for over 150 years . However, the inherently interdisciplinary nature of terpenoid research -- spanning fields such as chemistry, pharmacology, and biology -- poses significant challenges in integrat ing and communicati ng domain - specific knowledge across disciplines . To bridge this gap, we present TeroSeek, first by systematically extracting key scientific data and findings from terpenoid - related literature pub lished over the past two decades to construct a cura ted knowledge base (KB), and then further develop ing an intelligent question - answering chatbot and web service powered by an AI - accelerated retrieval - augmented generation (RAG) framework . TeroSeek en able s rapid access to structured, high - quality information and accurately respon ds to a wide range of terpenoid - related queries, demonstrat ing superior performance over general - purpose large language models (LLMs) in various application scenarios . T here fore, we believe that TeroSeek serves as a powerful domain - specific expert model to support the multidisciplinary terpenoid research community . The TeroSeek web service is publicly accessible at http://teroseek.qmclab.com .