PANDORA: Phylogeny-Aware Network Discovery of Microbial Roles and Annotation

Objective

"Microbes are the hidden engines of our planet, yet the function of most of their genes remains a mystery (the ""microbial dark matter""). Our understanding is held back because we typically recover only incomplete genomes from nature, making traditional analysis methods unreliable. This project will build PANDORA, a powerful predictive tool designed to overcome this fundamental barrier. PANDORA uses a phylogenetically-aware machine learning approach, trained on thousands of complete genomes, to learn the complex rules of which genes work together. This unique method allows it to accurately predict gene function even in fragmented data, distinguishing a truly absent gene from one that is simply missing. The project will: 1. Create the first large-scale, open-access library of predictive models for the global scientific community. 2. Discover the roles of thousands of unknown genes, generating a public roadmap for future biological research. 3. Reconstruct a timeline of microbial evolution, revealing when key processes like respiration and antibiotic resistance first emerged on Earth. By turning genomic data into functional knowledge, PANDORA will accelerate discovery. Its applications will advance Europe's bioeconomy through the discovery of novel enzymes, support public health by identifying new targets to combat antimicrobial resistance, and improve climate models by clarifying the role of microbes in global nutrient cycles."