TeroACT: A terpenoid bioactivity landscape and discovery platform.

Area of Science:

• Pharmacology and Cheminformatics
• Computational Biology and Drug Discovery

Background:

• Terpenoids possess diverse biological activities with significant pharmacological potential.
• Data-driven deep learning models are vital for efficient feature representation and knowledge inference in modern drug discovery.
• A comprehensive, multi-dimensional database is needed to map terpenoid-bioactivity profiles and explore their uncharted potential.

Purpose of the Study:

• To construct a large-scale biological knowledge graph integrating terpenoids, targets, genes, and diseases.
• To develop AI-driven predictive models for disease association and compound-protein interactions.
• To create a user-friendly web platform (TeroACT) for accessing these resources and facilitating terpenoid bioactivity research.

Main Methods:

• Integrated diverse data types (compounds, proteins, genes, diseases) to build a biological knowledge graph.
• Developed network-based disease prediction and compound-protein interaction models.
• Deployed resources on the TeroACT web platform (http://terokit.qmclab.com/teroact/) for user access.
• Utilized in silico models for screening terpenoids against melanoma.
• Validated findings using in vitro and in vivo experimental models.

Main Results:

• Successfully constructed a large-scale biological knowledge graph of terpenoids and related entities.
• Developed and deployed TeroACT, a web platform featuring predictive models for drug discovery.
• Identified mollugin and columbianadin with validated anti-migration and anti-proliferative effects in melanoma via in silico, in vitro, and in vivo studies.
• Discovered multiple terpenoids with anti-inflammatory properties through integrated computational and experimental approaches.

Conclusions:

• TeroACT provides a comprehensive data resource and AI-driven tools, bridging the gap in terpenoid bioactivity research.
• The platform facilitates efficient exploration of terpenoid bioactivity and supports the discovery of novel therapeutic agents.
• This integrated approach accelerates the identification and validation of drug candidates from terpenoid compounds.