From patterns to prediction: machine learning and antifungal resistance biomarker discovery.

Abstract

Fungal pathogens significantly impact human health, agriculture, and ecosystems, with infections leading to high morbidity and mortality, especially among immunocompromised individuals. The increasing prevalence of antifungal resistance (AFR) exacerbates these challenges, limiting the effectiveness of current treatments. Identifying robust biomarkers associated AFR could accelerate targeted diagnosis, shorten decision time for treatment strategies, and improve patient health. This paper examines traditional avenues of AFR biomarker detection, contrasting them with the increasingly effective role of machine learning (ML) in advancing diagnostic and therapeutic strategies. The integration of ML with technologies such as mass spectrometry, molecular dynamics, and various omics-based approaches often results in the discovery of diverse and novel resistance biomarkers. ML's capability to analyse complex data patterns enhances the identification of resistance biomarkers and potential drug targets, offering innovative solutions to AFR management. This paper highlights the importance of interdisciplinary approaches and continued innovation in leveraging ML to combat AFR, aiming for more effective and targeted treatments for fungal infections.