Advancing Natural Product Discovery: A Structure-Oriented Fractions Screening Platform for Compound Annotation and...

Area of Science:

• Natural Product Chemistry
• Cheminformatics
• Machine Learning in Drug Discovery

Background:

• Natural product discovery is often limited by the absence of tools that can integrate untargeted nuclear magnetic resonance (NMR) and mass spectrometry (MS) data on a large scale.
• Developing efficient methods for isolating and characterizing novel compounds from natural sources is crucial for identifying new therapeutics.

Purpose of the Study:

• To introduce and apply the novel Structure-Oriented Fractions Screening Platform (SFSP), an innovative NMR/MS-based machine learning tool.
• To demonstrate the capability of SFSP in enabling functional-group-guided fractionation and accelerating the discovery of undescribed natural products.

Main Methods:

• Application of the SFSP tool to analyze the extract of a marine fungus, *Aspergillus* sp. GE2-6.
• Utilizing SFSP for functional-group-guided fractionation based on integrated NMR and MS data.
• Isolation and characterization of flavipidin derivatives and phenalenone analogues.

Main Results:

• SFSP facilitated the isolation of 24 flavipidin derivatives and five phenalenone analogues, including 27 previously undescribed compounds.
• Structural elucidation revealed novel isomeric derivatives of flavipidins with unique ring fusions.
• Several isolated compounds, including flavipidin A and asperphenalenones, demonstrated significant anti-influenza and anti-HIV activities.

Conclusions:

• The SFSP tool effectively streamlines natural product isolation by integrating NMR and MS data.
• This approach significantly accelerates the discovery and characterization of novel chemical entities from complex natural sources.
• The identified compounds represent a valuable source for potential therapeutic agents, particularly against influenza and HIV.