Data-driven identification of potential multi-target antifungal agents for deep mycoses via machine learning and structural modeling

Introduction: Deep mycoses are severe fungal infections affecting subcutaneous or deeper tissues, often with high mortality rates, with opportunistic mycoses affecting over 3 million individuals globally. Additionally, neglected fungal pathogens such as Mucorales, Sporothrix spp., and Histoplasma spp. affect vulnerable populations, and lack adequate reporting systems. The treatment of these diseases is increasingly compromised by emerging drug resistance, highlighting the need for novel therapeutic strategies.
Methods: This study employed a sequential virtual screening pipeline to screen libraries of natural products and investigational drugs. It integrated five biologically grouped QSAR models built using the eXtreme Gradient Boosting (XGBoost) algorithm, followed by pharmacokinetic profiling. Prioritized hits underwent multi-target molecular docking against eight targets and were validated through molecular dynamics (MD) simulations to assess thermodynamic stability.
Results: The virtual screening and subsequent validations identified Citreoindole and TimTec1_006771, from the COCONUT database, as lead candidates. TimTec1_006771 emerged as the superior hit, exhibiting favorable interactions across multiple targets, including squalene epoxidase, squalene synthase, and exo-β-(1,3)-glucanase.
Discussion: Molecular dynamics results suggests that TimTec1_006771 is capable of multi-target inhibition, simultaneously disrupting membrane ergosterol biosynthesis and fungal cell wall remodeling. This suggests a synergistic effect that increases the evolutionary barrier against resistance. Furthermore, these candidates occupy “beyond Rule of Five” (bRo5) chemical space, suggesting that larger, rigid scaffolds may improve anchoring within diverse binding sites.
Conclusion: This research provides a computational framework for developing multi-target therapeutics against fungal pathogens. Future research should prioritize in vitro enzymatic and antifungal validation of these compounds, specially against resistant strains.