Leveraging AI to predict strain-specific susceptibility of Candida pathogenic yeast to novel antifungal "resistance breakers"

Systemic fungal infections caused by Candida yeast species are a growing public health concern, with high mortality rates and rising resistance to conventional antifungals. Emerging evidence indicates that environmental exposure to agricultural fungicides contributes to resistance in clinical strains, creating a critical link between crop protection practices and human health. Addressing this challenge requires a systems-level understanding of how agricultural practices shape fungal evolution and therapeutic outcomes.

This project aims to leverage artificial intelligence (AI) and genomics to transform our understanding of these complex interactions and inform more sustainable agricultural and clinical strategies. We will generate a comprehensive dataset comprising over 400 Candida isolates sourced from both clinical infections and agricultural environments in Northern Ireland. Each isolate will be fully sequenced, and its resistance profile to conventional antifungals and novel “resistance breaker” compounds will be experimentally characterized.

Applying transfer learning and deep learning, we will model high-dimensional genomic data to: (1) identify genetic signatures associated with fungicide-driven resistance, (2) predict strain-specific susceptibility to resistance breakers, and (3) uncover biomarkers capable of guiding both clinical therapy and agricultural interventions. AI-driven predictions will be validated through cross-validation and held-out datasets, with checks for class imbalance and consistency across clinical and agricultural samples, to ensure robustness and reproducibility.

The project will generate multiple tangible outputs, including a curated genomic dataset, predictive AI models, validated biomarkers, and evidence-based recommendations for sustainable fungicide use. Risk maps highlighting environmental hotspots for resistance emergence will inform crop management policies, while engagement with farmers and stakeholders will support practical implementation.

By integrating genomics, AI, and a One Health perspective, this project provides a blueprint for sustainable agri-food systems where crop protection practices are optimized without compromising public health. It exemplifies how AI can drive innovation in sustainable food systems, linking agricultural practices, environmental stewardship, and human health in a single, actionable framework.