READRetro web: A user-friendly platform for predicting plant natural product biosynthesis.

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Drug Discovery

Background:

• Natural products (NPs) are crucial bioactive molecules for pharmaceutical research.
• Large-scale NP production faces challenges in availability and scalability.
• Deep learning models accelerate NP discovery but often require significant computational expertise.

Purpose of the Study:

• To develop a user-friendly web platform for retrosynthesis prediction.
• To make advanced machine learning models for NP discovery accessible to researchers.
• To facilitate the design of natural product retrosynthetic pathways.

Main Methods:

• Integration of the READRetro machine learning model into a web platform.
• Utilization of modern web technologies for a fast and responsive user experience.
• Development of an accessible interface for retrosynthesis prediction.

Main Results:

• READRetro Web provides high predictive accuracy and computational efficiency.
• The platform offers well-organized results for designing NP retrosynthetic pathways.
• The web tool simplifies complex computational workflows for researchers.

Conclusions:

• READRetro Web democratizes access to advanced ML-driven retrosynthesis.
• The platform bridges the gap between computational tools and practical research.
• It empowers a wider range of researchers in natural product drug discovery.