Machine Learning-Driven Discovery and Database of Cyanobacteria Bioactive Compounds: A Resource for Therapeutics and...

Area of Science:

• Biotechnology and Bioinformatics
• Drug Discovery
• Environmental Science

Background:

• Cyanobacteria produce bioactive compounds with potential applications in medicine and environmental cleanup.
• A comprehensive resource is needed to catalog these compounds and their properties for research and industry.
• Existing data on cyanobacteria bioactive compounds is fragmented, hindering efficient exploration.

Purpose of the Study:

• To develop a searchable, curated database of cyanobacteria bioactive compounds.
• To build a machine learning model for predicting the targets of novel cyanobacteria molecules.
• To facilitate the identification of compounds for therapeutic and bioremediation applications.

Main Methods:

• A Python protocol extracted 3431 compounds and 373 protein targets.
• Chemical descriptors were calculated using PaDEL-descriptor, Mordred, and Drugtax software.
• Machine learning models were employed to predict compound-target interactions.

Main Results:

• A FAIR-compliant database of cyanobacteria bioactive compounds was established.
• A machine learning model was developed to predict molecular targets for drug discovery and bioremediation.
• The study identified promising protein targets for human therapeutics and environmental pollutant degradation.

Conclusions:

• The integrated database and machine learning model offer a powerful tool for researchers in pharmaceuticals and bioremediation.
• This resource accelerates the discovery of cyanobacteria-derived solutions for health and environmental challenges.
• The study highlights the potential of data science and machine learning in harnessing cyanobacteria for sustainable development.